Shifting human needs throughout the pandemic

Many of the existing studies and datasets of the Covid-19 pandemic focus on the biomedical and epidemiological aspects of the case and fatality rates, including efforts in detection or intervention and vaccine design. However, it’s important to keep in mind that, although we may not all be infected by the coronavirus, we are all affected by the coronavirus with challenges spanning societal, economic, and psychosocial realms.

To identify and understand the multi-level system of human needs and the impact of the pandemic and its associated policy decision on these needs, we developed a computational framework to characterize human needs. Our need categories were largely inspired by Maslow’s hierarchy needs, and we utilize difference-in-differences approach to account for various confounding factors.

At a glance, all needs are elevated during the pandemic. However, when looking at individual need subcategories, we find that basic needs (physiological and safety needs) were heightened while higher needs (self-actualization and cognitive needs) were subdued.

But the pandemic has disproportionately and negatively impacted socioeconomically and environmentally disadvantaged subpopulations.

Public health policies and interventions aimed at addressing the underlying socioeconomic and environmental conditions have shown to improve health outcomes. In recent decades, digital access has also gained attention as another factor mediating the health outcomes, as individuals utilize the Internet to seek health information as well as directly access healthcare services (i.e., telehealth, online pharmacy). Beyond health, digital resources play an important role in facilitating the fulfillment of a whole spectrum of human needs, some of which were exacerbated by mandated lockdowns.

Unfortunately, disparities in digital access also exist across the socioeconomic and environmental factors.

Health.

Given higher rate of pre-existing health conditions, existing disparities in healthcare access, and higher COVID-19 case and mortality rates, low SES subpopulations should be seeking more information about their health conditions. Instead, we find that ZIP codes with lower income show less change in health condition queries. For example, a ZIP code that was making 1,000 health condition queries a month before the pandemic makes about 10,000 such queries a month during the pandemic, but a similar ZIP code would only make about 8,000 such queries a month if that ZIP code had lower income.

Economy.

The COVID-19 pandemic has brought over 35 times increase in unemployment claims during the first month which hit the most vulnerable subpopulations the hardest. We find that stimulus related queries increased by over 15,000% in April on average, but ZIP codes with lower income show over 5000 percentage point more change in stimulate related queries at 5 weeks after the pandemic is declared. That means a ZIP code making 1,000 stimulus related queries a year before the pandemic makes 1,000 such queries every other day during the pandemic, but only one every three days if that ZIP code had lower income.

Learning.

At the beginning of the pandemic, many schools were closed while they were trying to figure out how to faciliate remote learning. Many schools and parents were forced to utilize online resources to educate their children. When we examine search queries that result in visits to free online learning resources (e.g., coursera.org, khanacademy.org), we find that ZIP codes with lower income and more Hispanic population exhibit close to half of the increase in those queries by their counterpart groups. ZIP codes with more Black population and high population density also show similar trend.