The Fragile Architecture of the EU’s Corporate Sustainability Due Diligence Directive

In December 2025, the European Union adopted the Omnibus I directive, introducing significant changes to the Corporate Sustainability Due Diligence Directive (CS3D). The amendments scaled back several provisions, deharmonised civil liability rules, and deleted climate transition plan obligations. This legislative revision reflects the ongoing political turbulence surrounding one of the EU’s most ambitious regulatory experiments: transforming large corporations into gatekeepers of sustainable global supply chains.

The CS3D has been controversial since its inception, facing substantial opposition from business groups concerned about compliance costs and extraterritorial reach. Yet much of the debate has focused on political and legal considerations, with less attention paid to the economic incentives created by the directive. In a recent paper, we analyse the economic logic behind the CS3D and apply well-established frameworks from law and economics scholarship to reveal several overlooked problems. Our analysis suggests that the directive’s mechanisms may prove inefficient and that the Omnibus amendments do not necessarily resolve these fundamental issues.

In principle, the CS3D could be justified as a tool for correcting market failures. However, the mechanisms employed may well create new distortions. The CS3D’s due diligence requirements function as a sort of indirect ‘Pigouvian tax’: they make trade more expensive by introducing costs of verification and monitoring throughout supply chains. Yet unlike a well-designed Pigouvian tax that targets harmful activities precisely, the CS3D imposes costs indiscriminately across all industries. The Omnibus amendments partially addressed this by requiring firms to adopt a risk-based approach and prioritise larger or more likely harms, but the correction is incomplete: the risk classification still occurs at the firm level and at the firms’ discretion.

A follow-up question is: who bears the cost of the implicit tax? We address this by turning to the Coase Theorem: if transaction costs between actors in the supply chain are sufficiently low, the actual bearer of the cost will be the one who can do so most cheaply (the least-cost avoider). Thus, if transaction costs were low, the directive’s assignment of responsibility to EU firms would not matter much for efficiency: firms would negotiate with their suppliers, rolling down costs so that the best-positioned party would ultimately prevent harm. However, transaction costs in global supply chains may be quite high due to various factors, including information asymmetries, cultural and language barriers, geographic distance, and legal uncertainties. When transaction costs are high, the initial allocation of responsibility matters enormously: it achieves an efficient allocation only if EU firms happen to be the least cost avoider. Our paper uses a numerical example to show how three inefficient outcomes become likely: (i) EU corporations might (inefficiently) implement expensive prevention measures themselves; (ii) they might exit the market entirely if combined costs exceed profits; or (iii) they might switch to developed-country suppliers with higher production costs but lower monitoring costs, even when the original suppliers could have prevented harm more efficiently.

While the CS3D contains some provisions aimed at reducing transaction costs, such as resource sharing within business groups, digital tools, model clauses, and help desks, these measures are unlikely to fully bridge the gap. The Omnibus further complicates matters. On the one hand, it asks firms to conduct in-depth assessments only if a preliminary ‘scoping exercise’ indicates a need to do so. This can reduce transaction costs, as not all transactions require full inspection. On the other hand, it permits companies to request information solely when doing so is ‘necessary,’ which can become a source of dispute and uncertainty, possibly leading to higher transaction costs. 

Our analysis of the enforcement structure reveals additional concerns. Drawing on Steven Shavell’s seminal framework for understanding when regulation, liability, or their combination is optimal, we examine how the CS3D’s dual enforcement mechanism: administrative fines plus (deharmonised) civil liability, may create problematic incentives. When regulation and liability are intertwined such that compliance with one exempts from the other, two simultaneous deterrence failures can emerge. Firms with high compliance costs may be overdeterred, taking excessive precautions that cost more than the harms they prevent. Meanwhile, firms with low compliance costs face underdeterrence: they can ‘just comply’ with the regulatory standard rather than taking additional but efficient precautions, since compliance exempts them from both fines and liability.

The Omnibus amendments, rather than resolving this tension, may have exacerbated it. The deletion of the harmonised liability regime means civil liability is now defined by each Member State’s national law, creating a patchwork where compliance may shield a firm in one jurisdiction but not another. The removal of ‘overriding mandatory provision,’ which removes the obligation to ensure that only EU applies, adds foreign law to the mix. New safe harbours, in which firms prioritising the most severe impacts cannot be penalised for failing to address less significant ones, create compliance-as-defence structures whose scope remains uncertain across jurisdictions. The cap on maximum fines (3% of net worldwide turnover) while leaving civil liability unbounded produces a systematic asymmetry: regulatory exposure becomes calculable, tempting some firms to treat fines as a cost of doing business, while civil liability remains unpredictable.

Beyond these incentive problems, we identify several unintended consequences. Perhaps most troubling is a statistical discrimination dynamic: when EU corporations cannot efficiently monitor suppliers in developing countries, they may avoid these suppliers entirely. The Omnibus’s scoping exercise, based on geography and enforcement levels, may amplify this by encouraging firms to follow rules of thumb that lead to avoiding suspicious countries altogether. Additional concerns include regressive price increases that disproportionately burden low-income consumers, economies of scale in compliance that favour very large firms and increase market concentration, and the potential for tacit collusion among competitors who coordinate their monitoring processes.

These findings expose what we call the ‘fragile architecture’ of the directive. The CS3D represents a well-intentioned but potentially flawed regulatory framework whose design creates mixed incentive structures that may undermine its stated objectives. The Omnibus amendments, adopted amid intense political pressure, do not address these fundamental economic problems, rather, they merely relocate them. The goal should be regulatory design that achieves genuine environmental and human rights improvements through efficient incentives, without inadvertently harming the populations it aims to protect.

The authors’ full paper, ‘The Corporate Sustainability Due Diligence Directive: A Law & Economics Evaluation’, can be found here. 

Hadar Yoana Jabotinsky is the Founder of the Hadar Jabotinsky Center for Interdisciplinary Research of Financial Markets, Crises and Technology.

Roee Sarel is a Professor of Private Law and Law & Economics at the Institute of Law and Economics, University of Hamburg.