Working Paper: NBER ID: w31919
Authors: Amalia R. Miller; Kamalini Ramdas; Alp Sungu
Abstract: We measure the impact of the initial Indian national COVID-19 lockdown on digital activity using browser histories of 1,094 individuals, spanning over 31.5 million website visits on computers and mobile devices. Reflecting the predicted increase in the value of online activity, both men and women in our sample dramatically increased their internet browsing during the lockdown. However, men’s browsing increased by significantly more, causing gender gaps overall and in key browsing categories, and in browsing on mobile devices. Our browser data showed significant relative reductions in women’s online job search, corroborated in aggregate data obtained from a major Indian online job platform, indicating potentially persistent harms to women’s employment. Consistent with increased childcare obligations driving the observed gender gaps, we find that gaps were greatest among parents. Men and women in our sample had similar browsing levels and trends pre-pandemic, which diverged during the lockdown. Our primary findings therefore shed new light on determinants of digital time use, while also highlighting the importance of considering both extensive and intensive margins of digital activity to track the digital divide. In our secondary analysis of time devoted to childcare, we find conflicting survey responses between fathers (who report an increase relative to mothers) and mothers (who report no such increase). While our data cannot directly resolve this conflict, they do show fathers having larger increases in time spent online, with no relative increase in childcare-related browsing. This secondary result demonstrates the value of complementing survey data with digital trace data.
Keywords: computer and mobile browsing activity; COVID-19 lockdown; digital time use; gender; digital divide
JEL Codes: I18; J13; J16; J22; J24; K39; O15; O33
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Lockdown (K42) | Increase in internet browsing for both men and women (L86) |
| Lockdown (K42) | Increase in men's browsing activity (J29) |
| Lockdown (K42) | Increase in women's browsing activity (J16) |
| Lockdown (K42) | Decrease in women's online job search activity (J29) |
| Increased childcare responsibilities (J13) | Decrease in women's online job search activity (J29) |
| Lockdown (K42) | Gender gaps in digital activity (J16) |