Taxation and Supplier Networks: Evidence from India

Abstract

Do tax systems distort firm-to-firm trade? This paper considers the effect of tax policy on supplier networks in a large developing economy, the state of West Bengal in India. Using administrative panel data on firms, including transaction data for 4.8 million supplier-client pairs, we first document substantial segmentation of supply chains between firms paying Value-Added Taxes (VAT) and non-VAT-paying firms. We then develop a model of firms' sourcing and tax decisions within supply chains to understand the mechanisms through which tax policy interacts with supply networks. The model predicts partial segmentation in equilibrium because of both supply-chain distortions (taxes affect how much firms trade with each other) and strategic complementarities in firms' tax choices. Finally, we test the model's predictions using variations over time within-firm and within supplier-client pairs. We find that the tax system distorts firms' sourcing decisions, and suggestive evidence of strategic complementarities in firms' tax choices within supplier networks.

Does going cashless make you tax-rich? Evidence from India’s demonetization experiment

“…We estimate that, in aggregate, firms report total sales to the tax authorities that are substantially higher than the total electronic sales in their areas (see Appendix A.1.6). This is consistent with evidence in Gadenne et al (2021) that firms report sales that are much higher than their total third-party reported sales and suggests firms pay taxes on more than just their electronic sales. We therefore do not expect firms to simply increase their reported sales by the amount of extra electronic transactions they make during demonetization.…”

Relational Contracts and Development

“…8 Readers are referred to Martin et al (2020) for the case of France, Macchiavello (2010) for Chilean wine exports, Macchiavello & Morjaria (2015) for Kenyan cut flower exports, and Cajal-Grossi et al (2022) for exports of Bangladeshi garments. 9 VAT data have been used in Belgium (Dhyne et al, 2015), Chile (Huneeus, 2018), Costa Rica (Alfaro-Urena et al, 2019, the Dominican Republic (Cardoza et al, 2020), Ecuador (Adao et al 2020, Turkey (Demir et al 2021), Uganda and Rwanda (Spray 2021), and India (Gadenne et al, 2019). Whereas these data were originally used in public finance, recent studies have focused on firm-to-firm networks, diffusion of shocks, and spillovers.…”

COVID-19 and taxes: Policies for the post-pandemic recovery

“…75 Similarly, compliance shocks had a multiplier effect across supply chains in West Bengal, India. 76 Because complying firms prefer trading and dealing with other firms of the same type, formalisation can potentially spread through the production network.…”