Research database

This research proposal addresses the environmental impact of economic downturns, taking the COVID-19 shock as a natural experiment. While the COVID-19-imposed lockdowns had important short-term environmental effects (Elsaid et al. 2021), the implications of the COVID-incepted economic crisis may bear long run sustainability implications (Hepburn et al., 2020; OECD, 2021). Indeed, the shrinking margin of firms decreases the resource buffer and increases the incentives to cut on input costs, which may lead firms to reallocate their input portfolio towards lower-quality inputs, with potential implications for their longer-term sustainability (Antweiler et al., 2001; Copeland and Taylor, 2004). If ``greener’’ inputs are more expensive than polluting inputs, the production technology of affected firms may have become more polluting (e.g., Frankel & Rose, 2005; Egger et al., 2021). At the same time, the sudden halt imposed by the pandemic to economic activities may have represented an opportunity to revise the production technology and to reduce inefficiencies, including the use of polluting inputs, in line with classical arguments of the Porter Hypothesis (PH; Porter and Van den Linde, 1995; see also e.g., Devece et al. 2016). In particular, the increased transaction costs arising from the pandemic may have reduced the incentives for firms to vertically dis-integrate, reduced the incentives to offshore production and possibly even to outsource domestically. The environmental impacts of these changes may be relevant if the pandemic represented an opportunity to invert decisions initially motivated by ``pollution haven’’ considerations (Levinson and Taylor, 2008), bringing production back to countries with stricter environmental regulations and reducing the environmental costs of cross international transportation. Empirically, this proposal exploits the COVID-19 pandemic as a shock that exogenously imposed a temporary halt to production depending on whether the firm’s industry was classified as ``essential’’ vs. ``non-essential’’. The exogenous assignment of firms to essential/non-essential sectors can be exploited in a diff-in-diff approach (cfr. Di Porto et al., 2022, Goodman-Bacon and Marcus 2020). Firms more strongly affected by temporary closures had to face production slowdowns, value-chain disruptions, and revenue erosion, hence will have stronger reasons to revise their production strategies. To measure input adoption, we will critically rely on the availability of unique population data on Italian firms’ international imports. Under some circumstances, imports may represent a valid measure of input adoption (Brunel, 2019). Drawing on detailed information about the value and the sectoral and product classification of imports, we will be able to derive a fairly accurate measure of “greenness” of the imports and to verify the underlying assumption that ``green’’ inputs are more expensive than ``dirty’’ ones and study whether Italian firms reacted to the COVID-19 shock by adjusting their input choices. As such, the results will not only inform about the likely environmental impacts of economic downturns, but also speak to the broader debate about the trade-off between ``growing and greening” the economy (e.g., Brandi et al, 2020).