Paul Stoddart

I am just back from Milan, where I joined a policy panel at the European Trade Study Group (ETSG) 2025 conference. I made three main points: 1) The trade-weighted average US import tariff has risen from 2.5% to 18.8% this year, which will lead to a significant slowdown in global trade growth. 2) The US, however, accounts for only about 14% of world merchandise imports, and roughly 72% of world trade continues to be conducted directly under the WTO’s MFN tariffs. 3) It is therefore essential that economies do not undermine rules-based trade with the rest of the world when pursuing trade deals with the US. Concretely, this means upholding MFN and avoiding arrangements that disadvantage third countries. It is also critical to manage trade diversion effects cooperatively, recognizing that it typically is a two-way street. UZH Department of Economics UBS Center for Economics in Society

331

11 Comments

Ten essential principles for economic policymakers outlined by econ-professor-turned-politician Andrew Leigh MP: 1) Focus on well-being, not just money 2) Think comparative advantage 3) Ignore sunk costs 4) Optimize, subject to constraints 5) Use the best evidence 6) Consider expected value 7) Think in magnitudes, not just signs 8) Ask: what is the market failure? 9) Consider the distributional consequences 10) Incentives matter Read the full article here: Andrew Leigh (2024), Ten Lessons for Economic Policymakers, Economic Papers: https://lnkd.in/ed8wQwjt https://lnkd.in/e-Z-YqYh (ungated)

211

4 Comments

With the fall teaching season approaching, I wanted to reshare my working paper: "Understanding Market Power". This paper is a comprehensive guide to the tricky yet essential economics behind antitrust law. It is intended to be clear and broad, covering key law & econ topics in the US and the EU. I welcome your feedback in anticipation of next year's version. Link to paper https://lnkd.in/eY4bzJSN #Antitrust #Economics #CompetitionPolicy #Law

371

6 Comments

My hot take on the European Commission's data disclosure template for general-purpose AI models. First and foremost, a core requirement remains unclear. Model providers must disclose "the top 10% of all domain names determined by the size of the content scraped (in a representative manner across all modalities where applicable)." But what exactly does a "representative manner" mean here? Is the 10% calculated per modality (text, audio, video, etc.) or across all modalities? Needless to say, this matters a lot. How one calculates the "size of the content scraped" is also a crucial question. Are we talking about raw file size (e.g. megabytes) or something more meaningful like the number of extractable tokens? The latter would provide a fairer picture, otherwise, video or high-resolution image data could skew the dataset and misrepresent content importance. On the scope, the Commission seems to interpret the concept of 'training' to go beyond the placement on the market for models that continue learning once deployed. The explanatory notice also clarifies that the template will apply to downstream players that modify a model so much that it becomes a new one (i.e. passing the one-third threshold), but only concerning the data added with the modification. The template could become a powerful tool in copyright enforcement. Rights holders may now trace whether their content was used in training. AI firms must explain crawler behavior, including how they treat paywalled content and comply with the Text and Data Mining (TDM) exception in the Copyright Directive. Finally, will the template itself be litigated? This is a mandatory tool, so it could be challenged in court. The Commission claims it balances transparency with protection of trade secrets, but will AI developers agree? Some may argue it oversteps the AI Act, especially if they believe it compels disclosure of sensitive business information or goes beyond the 'narrative' format. Plenty of questions left unanswered for now. In the meantime, here is my write-up for MLex 👇

125

6 Comments

🚨 New Publication Alert 🚨 I’m excited to share a new paper co-authored with Tobias Kawalec just released as part of the World Bank’s Policy Research Working Paper Series. 💡 What we do: We use state-of-the-art econometric techniques (Local Projection–Instrumental Variable) to estimate fiscal multipliers for all Gulf Cooperation Council countries (Bahrain, Kuwait, Oman, Qatar, Saudi Arabia and UAE). In other words, we use global hydrocarbon demand shocks to study fiscal multipliers in resource-rich economies. 📊 What we find: -Short-run fiscal expenditure multipliers lie in the range of 0.1–0.4. -Multipliers rise toward the upper end during recessions, suggesting fiscal policy is especially effective in Gulf countries during downturns. 🙏 A huge thanks to Safaa El-Kogali, Sandeep Mahajan, Nadir Mohammed, Jasmin Chakeri, Eric Le Borgne and Hoda Youssef for their guidance and support. And to Wael Mansour, Desislava Nikolova and Dániel Prinz for their excellent comments. 📄 Read the paper here: https://lnkd.in/gtQwY7-S

156

3 Comments