Estimating the Impact of Digitalization on Firm Performance During COVID-19

Abstract Using the coronavirus pandemic (COVID-19) as a natural experiment, recent studies have shown that digitalization has been key for improved firm performance. While earlier studies mostly use the canonical binary difference-in-difference design, recent advances have shown that this may lead to biased results when the digitalization treatment variable is continuous (treatment heterogeneity). In this paper we show that digitalization, measured in terms of online sales, has indeed been a significant driver of firm performance during the coronavirus pandemic (COVID-19) also when considering treatment heterogeneity. Using World Bank Enterprise Surveys data for ten Western European countries shows that initially fully digitalized firms with 100% of sales online experienced 10%-point higher firm-level sales growth during the pandemic when average firm sales contracted by 8.7%. Our results are robust to controlling for unobserved and observed firm heterogeneity, treatment heterogeneity and sample selectivity, while the common trend assumption is validated and further robustness is confirmed using a placebo test. This study contributes to the literature by providing a rigorous estimate of the importance of digitalization for firm performance within the context of a natural experiment.