The
outbreak of COVID-19 has put the global community of evaluators at an interesting,
almost paradoxical, juncture. While the need for data and insights is most
critical at this point to respond to the public health crisis, collecting this
data is going to be now, and henceforth, more challenging than ever.
The
article ‘Rewiring How We
Measure Impact in the Post-Covid-19 World’ co-authored by Veronica Olazabal, Michael
Bamberger, and Peter York puts several of our concerns as evaluators into perspective.
In addition to the challenges that COVID-19 will potentially create for
evaluators, the article stresses how Data
Science will be instrumental in how the development and impact sector responds
to the need for evidence.
With
social distancing norms dictating human interactions until the pandemic is
effectively controlled, in-person data collection is going to be hindered as
the article correctly establishes. Yet, the need for real-time analytics from
data is going to be even more critical not just in the public health domain but
across parameters in order to rebuild the damage COVID-19 is inflicting on the socio-economic
fabric of countries universally. Granular, localized, and verified data will be
required by decision-makers – public, private, or philanthropic – in order to
carry forward fast and effective impact at scale.
Data
Science for development and social impact has been steadily gaining buy-in from
governments, philanthropies, and relevant organizations across the board for a
while now. Data Science technologies offer several advantages. Diverse datasets
and sources can be integrated, and sampling biases can also be removed as big
data covers large populations. Data Science also allows us to study patterns in
behavior over long periods of time as well as predict future trends, an aspect
that private sector organizations have been successfully leveraging.
The current COVID-19
situation will push the sector to treat Data Science not solely as an
optimization tool, but as a critical necessity for social justice and impact to
be delivered. This holds particularly true in Global South contexts where
damage caused by the pandemic can be exponentially catastrophic.
India will also have to channelize Data Science technologies in
order to contain the outbreak and to recover from the damage COVID-19 is
inflicting.
Fortunately, discourse is already underway in both public and private domains
on how big data and AI can be utilized. However, translating this on the ground
will be challenging for India owing to multiple factors. Not only is the architecture
to harness data on the ground weak, the culture for data usage and appreciation
is still nascent.
With
the contextual and behavioral complexities and diversity of the Indian
population, relying on big data sources such as social media, IoT, mobile phone
and geospatial data alone will be thoroughly insufficient. A large proportion
of this big data is harnessed from the supply-side, i.e., from the point of
sale. It may not reflect the actual behavior of users who are on the demand-side.
Unless big data can be validated and backed up with socio-economic and
demographic variables, it will be of limited use to policymakers and
implementors. Further, despite digital penetration in India being relatively
high and digital payments and media usage increasing in rural areas, the modes of
tapping big data is still relatively small for majority of the population. For
remote, marginalized populations, stray digital activity and behavior is
incapable of reflecting any valid or true patterns of human, demand-side
behavior.
What India
needs in current circumstances is a means to validate analytical insights
derived from Data Science technologies, and to optimize the already existing data
that exists at the grassroots that is yet to be tapped for its complete
potential.
Since
in-person data collection will be hindered, remote methods such as telephonic and video calling
technologies must be explored. Mobile
and high-speed internet penetration in India will enable this to a large extent
in India. Layering this with big data insights can overcome validity concerns
and offer holistic, 360-degree insights on human behavior patterns which could
also be used for forecasting activities.
We
will also have to inspect our public systems for the data reserves that already
exist. Several grassroots and local entities house rich primary, demand-side
data that can be transformative. For example, frontline community health
workers are well-trained in collecting medical and health data, several of who use
technology-enabled devices to track records. Tapping into this to maximize the
insights derived can prove immensely valuable.
Navigating
contextual constraints with technology and by leveraging existing
infrastructures is the answer to how evaluators in India will respond to the
need for data in the post-COVID-19 era. Data
Science will become indispensable to how the development sector advances, and
innovation will help overcome some of the challenges it poses in the Indian
context.
Swapnil Shekhar - Director and Co-founder, Sambodhi
Kaamila Patherya - Program Manager, Sambodhi
