As the U.S. and international climate debate gears up, we are getting a lot of questions about how WTO compliant is this or that climate policy. My colleague Steve Charnovitz and two co-authors recently released a book called Global Warming and the World Trading System that goes into some depth on this topic.
I hope to release a response to this report in the coming weeks or months, but Steve and his coauthors have an extremely useful appendix that summarizes key WTO cases with relevance for environmental protection. Drawing from this, as well as this win-loss chart and our recent report on the WTO compatibility of Obama's green jobs plans, I conclude that the attempts to defend environmental and other public-interest policies at the WTO have failed most of the time.
By way of a background, the WTO's General Agreement on Tariffs and Trade (GATT) prohibits discrimination against foreign products (even some measures not intended to discriminate), and makes difficult all sorts of other environmental policy implementation besides. That grouplet of trade lawyers that claim that the WTO doesn't represent the most significant international legal tripwire against environmental protection rely on GATT's Article 20 exceptions, which read:
(d) necessary to secure compliance with laws or regulations which are not inconsistent with the provisions of this Agreement, including those relating to customs enforcement, the enforcement of monopolies operated under paragraph 4 of Article II and Article XVII, the protection of patents, trade marks and copyrights, and the prevention of deceptive practices;...
(g) relating to the conservation of exhaustible natural resources if such measures are made effective in conjunction with restrictions on domestic production or consumption;
As you can see, defending an environmental measure can be cumbersome. There's that pesky "requirement that such measures are not applied in a manner which would constitute a means of arbitrary or unjustifiable discrimination between countries where the same conditions prevail, or a disguised restriction on international trade." Then there's also that funny word "necessary." Is the environment "necessary"? Is my awful plaid shirt that my dad gave me that I am offending my co-workers with "necessary"? What about that after-dinner ice cream? We focus on some stats on the necessary test in this post.
It turns out that meeting this "necessary" hurdle is very difficult. According to Joost Pauwelyn, GATT panels never deemed environmental or public-interest policies "necessary" prior to the establishment of the WTO. As we list in our report in footnote 130, there have been 11 post-95 WTO cases where Article 20 exceptions have been invoked.
- In In US - Shrimp 2006, EC - Trademarks 2003, EC - Tariff Preferences 2002, Canada- Wheat 2002, and Korea - Beef 1999, the necessity test was rejected. (This paper identified two pre-95 cases where necessity was invoked: 1989's US-Section 337 1989 and 1992's US - Beer. In both cases, the "necessity" defense was rejected. There were also two other GATT-era cases, US-Tuna Dolphin 1994, and Thailand-Cigarette 1990, where the necessity test was rejected. The Tuna-Dolphin case is one of the most infamous GATT cases, and is being started up again by Mexico now.)
- In Brazil-Tyres 2007, the WTO panel agreed with Brazil on necessity test, but not on the so-called chapeau requirement bolded above. And on EC-Asbestos 1998, the panel accepted EC's necessity defense. This is the first time this was true.
- The other cases, Argentina - Bovine Hides 1998, US-Gasoline 1995, Mexico - Soft Drink 2004, and Dominican Republic - Cigarettes 2003, relied on other exceptions and did not have to meet the necessity test.
So, out of 11 cases pre- and post-95 where the "necessary test" was invoked, it was only accepted twice. Of the 15 cases listed here, the overall exception was only accepted twice. So environmental and other GATT exceptions failed 80-87 percent of the time.
If you have comments, or know of other cases where the exceptions (and specifically ones where the "necessary" test is relevant), please let me know. This is what I'm cramming together from memory and these few sources.