I recently came across Nigerian-American artist Mimi Onuoha’s mixed media installation, The Library of Missing Datasets. The Library showcases a physical repository—a filing cabinet—of information that ought to exist, but due to cultural, social or other inherent biases or indifferences, does not. The incomplete list of missing datasets includes undocumented migrants currently incarcerated in the US and how often police arrest women for making false rape reports. This piece of work made me think about all the datasets that are currently blank spots in the response and recovery process to COVID-19 across Africa.

Since the start of the pandemic, there has been a deluge of online surveys on social media from entities such as the Uganda Investment Authority asking, “Has your business been affected by #COVID19?” The United Nations (UN) termed the rise in reported domestic violence cases  “the shadow pandemic.” The UN Population Fund estimated that for every three months of lockdown, 15 million more cases of domestic abuse could happen than would normally be expected, and millions of women could lose access to vital services, such as access to contraceptives. Even more devastating than these predictions, the Kenya Health Information survey reported close to 4,000 girls under the age of 19 “were impregnated” in a single county during the lockdown. About 200 of these girls were under the age of 14. Now, the very wording of these articles is problematic but provides an important lens to understanding exactly what framing of issues and data are missing and, likely, forever lost since the start of the pandemic. How many teenagers, who didn’t make it into the health survey, resorted to unsafe abortions? How many other children were sexually abused or raped?

Several governments across Africa have done a relatively good job in providing aggregate data on coronavirus cases, though often lacking in granularity, and with considerable challenges in contact tracing. Some countries still remain closed, while others lack structured plans to re-open or to provide income replacement or other forms of assistance to the many millions of citizens who have suffered and lost their livelihoods under lockdown conditions. In Uganda, based on the observed increase in maternal mortality in early March 2020 compared to the preceding 2019-2020 average, an excess 486 deaths are predicted for a six-month period, incurring 31,343 disability-adjusted life years (DALYs) lost. According to a recent Global Fund to Fight AIDS, TB and Malaria survey, 85% of HIV programs, 78% of TB programs, and 73% of malaria programs reported disruption to service delivery due to lockdowns, restrictions on transportation and COVID-19-related stigma. Estimates in some countries predict a 10%, 20%, and 36% increase in HIV, TB and malaria related deaths respectively over a five-year period.

However, these estimates cannot capture the true reality of the health outcomes across the continent. Data on morbidity and mortality from non-COVID causes but due to COVID lockdowns is largely missing—for example, women who died walking to hospital because public transport was cancelled. Data on the number of persons, mostly women, who suffer from Lupus-related pain due to shortages of hydroxychloroquine (HCQ) is missing. Data on the number of women who faced domestic abuse, and children who were sexually abused, is missing. Data on the response by governments to these women and children is, of course, also missing.

Furthermore, governments have largely ignored the needs of persons with disabilities, and other traditionally marginalized or minority groups, oftentimes resorting to violence to enforce restrictions and curfews. For example, from a global perspective, data released by the CDC in June showed that death rates among Black and Hispanic/Latino people are much higher than for white people, in all age categories. In Britain, Black citizens were up to two times more likely to die than white citizens. Other countries like France are still grappling with the ethics of collect data of race, ethnicity, and religion—affecting the country’s ability to protect the most vulnerable populations from the impact of the pandemic. However, data is crucial in guiding government decision-making, especially in the absence of an appropriate response to the needs of minority groups as articulated by civil society and local communities.

In Uganda, 19 LGBT+ persons were jailed for 50 days after their shelter was raided by police for violating social distancing rules, which banned gatherings of more than 10 people. In Nairobi, a 13-year old was killed by a stray bullet when police officers moved through his neighborhood in an attempt to enforce curfew. Also in Uganda, a young deaf and blind man was shot by a team of Local Defence Units in Agago District, in the northern region for breaking curfew, resulting in the amputation of his leg. How do caregivers for persons with disabilities practice social distancing? How do they get around when public transportation is restricted? How do persons with seeing or hearing impairments receive information on the pandemic?

In terms of educational outcomes, how do we even begin to measure the impact of COVID-19 on student achievement and learning losses? As of June 24, schools in up to 134 countries were either partially or completely closed, affecting more than 1 billion students. While wealthier, urban students can still tap into digital learning resources, their poorer and disconnected counterparts in both urban and rural settings will fall significantly behind. How many students are currently engaged in digital education? How many institutions are offering e-learning programs? What proportion of parents are able to contribute to home-schooling? We really don’t know.

Moving on to economic losses, the gaps in information are massive. Given that up to 85% of employment across Africa is informal, workers often lack any forms of social protection and are not adequately accounted for in research and planning. Market closures, disruptions to transportation, and reduced consumption have caused immeasurable suffering across the continent as a majority of businesses ground to a halt. How many businesses will never recover from the shock? How are people getting by with little to no support from the government?

An alarming assessment by the World Bank found that up to 65% of National Statistical offices were closed in the early months of the pandemic, resulting in major disruptions in the collection of essential and foundational data. These offices have been traditionally underfunded, used as propaganda machines or simply lack the skills, resources, and government commitments to support data ecosystems.

In an attempt to fill these gaps, companies such as TracFM have used SMS and radio polling to collect opinions from listeners, asking questions such as “What is your main concern during the coronavirus lockdown?” The World Bank, IMF, and McKinsey have all put out their own predictions on impact across different sectors. However, private companies, NGOs, and foreign consultants stepping in to measure and guesstimate is insufficient. Data is created, processed, and interpreted under unequal power relations and can reproduce the same oppressive and discriminatory norms that already exist in today’s societies. It’s important to understand the motives behind this data collection by foreign enterprises, to focus on supporting national statistical offices to bolster stronger systems and to ensure trust in the general public around the data in circulation.

African governments have been decisive since the outbreak. Safety net responses across the continent have included tax relief, cash transfers, food distribution, and utility bill freezes. It is understandable that collecting data during a pandemic can be a daunting exercise. However, this data is vital to understand where to prioritize recovery efforts. In the face of major economic losses and depleting tax revenues, how will social safety nets be implemented in a way that protects the most marginalized? Finally, how are governments applying gender-sensitive responses in post-crisis recovery plans—ones that address the structural inequalities that disproportionately impact women. Let’s make sure that these do not end up in the cabinet of missing datasets.