I enjoyed this recent JIEL article by Emily Lydgate, called "Is it Rational and Consistent? The Wto’s Surprising Role in Shaping Domestic Public Policy." Here is the abstract:
What makes regulation rational? And why is rationality important to an international tribunal? In the World Trade Organization (WTO) context, these questions have had significant implications for the public policy of its Member countries. The WTO Appellate Body’s emerging emphasis on means–ends rationality testing is based on the questionable premise that consistent regulation is non-discriminatory. It has led regulators, such as the EU, to defend—and probably even construct—complex regulation in a way that emphasizes conformity to one overarching policy objective. More surprisingly, the Appellate Body has re-cast itself as public policy watchdog, pointing out when governments do not appear to be committed to their cause. In response, governments have strengthened disputed regulation, rather than making it less trade-restrictive. This retreat to rationality can be seen as a result of a difficult challenge facing the Appellate Body: how to review national regulation without passing judgment on it. More specifically, the rationality test pays the price of the Appellate Body’s retreat from proportionality.
And here's an excerpt:
This rationality test raises questions about how WTO dispute settlement influences national public policy. Clearly the dispute settlement bodies critically review Member States’ regulation. This is to achieve a delimited goal: determining whether a domestic regulation disputed by another WTO Member State (or States) violates WTO law. In the context of the provisions cited above, the goal more specifically is to establish that the regulation in dispute does not discriminate against products from one Member State as compared to another’s (e.g. imported versus domestic products).
Yet a review of disputes in which this test has been applied reveals that the goal of regulatory review can become blurred due to spurious correlation between consistency, rationality, and non-discrimination. Establishing rationality as a goal suggests that regulation needs to be perfect and leads to the Panel and the Appellate Body imposing unrealistic standards. The rationality test rests in the grey area between the narrower objective of identifying the existence of discrimination and the broader aim of judging whether regulation is ‘good’; e.g. clear and effective. Although the WTO dispute settlement process may well result in ‘better’ regulation, this should be limited to a side-effect of eliminating discrimination (rather than the primary objective).
As I see it, there are four concepts that are used to review national regulation in various contexts: rationality, necessity, proportionality, and discrimination. It would be nice if there were a neat separation between them, but in practice they overlap quite a bit. Let me walk through them each briefly (but maybe not comprehensively).
Discrimination could refer to either a discriminatory effect or a discriminatory intent (or preferably, in my view, both). Here, we are talking about nationality-based discrimination, not some narrower concept.
In the WTO context, necessity involves a weighing and balancing of issues such as the contribution of the measure to its goal, the trade-restrictiveness of the measures, the importance of the goal, and the reasonable availability of less trade-restrictive measures (varying somewhat depending on the particular provision where it is applied).
Proportionality is complicated, but, as explained here, involves “suitability”, that is, whether "the means adopted by the government are rationally related to stated policy objectives"; and “necessity,” that is, the deployment of a “least-restrictive means” test. There is also “balancing in the strict sense,” where the benefits of the act are weighed against the costs incurred by infringement of the right.
Rationality refers to a means-ends relationship between the measure and its objective. On its face, this might seem less intrusive in relation to domestic regulation, but perhaps that depends on how you apply it. If you are asking governments to be "consistent," this might get pretty intrusive.
As is probably clear from the above explanations, things quickly get messy, because these concepts can overlap and work together. For example, what I think the Appellate Body has been doing in some non-discrimination contexts is reviewing whether there is a discriminatory effect, and if there is, using rationality as a way to determine whether there is a discriminatory intent. Is that the best say to evaluate intent? Personally, I prefer a broader, more comprehensive, objective examination of the intent of a measure, which I see reflected in paras. 56-75 of Chile - Alcohol. What exactly is the difference between the two approaches? It's hard to say how much practical impact there is. This really may be just be about the characterization of the approaches, and how clear the adjudicatory body chooses to be about what it is doing. In some cases, the emphasis seems to be on rationality itself, and the connection to the objective intent of the measure is downplayed. But in such cases, I'm not sure the results would differ if the approach were more expressly tied to the "design, structure, and architecture" of the measure.
Anyway, Emily's article provides an excuse for a consideration of all this, but the issue has been around since I got into IEL, and probably isn't going away soon.
