Someone asked me to speak to their organization about the problem of non-market policies and practices, and one of the points I was going to make was that while some people seem to think AD/CVD can be a useful tool to address at least a few of these policies and practices, the ability of AD/CVD to induce changes in these policies and practices is, as far as I can tell, close to zero. As I was thinking this point through, I came across the following quote from former U.S. Trade Rep. Robert Lighthizer in the latest issue of International Economy magazine:
In a series of decisions, the Appellate Body has struck down U.S. practices that are designed to control Chinese subsidies and illegal dumping. These cases have made it hard for the United States and other countries to counter industrial subsidies and other unfair actions.
Putting aside the characterization of the Appellate Body rulings and their impact on U.S. AD/CVD capabilities, his use of the words "control" and "counter" here may be helpful in understanding what might be effective against non-market policies and practices. I may be reading too much into his choice of words (to be honest, he might have just been trying to avoid repetition!), but "control" could be taken to mean a direct impact on the subsidies themselves and an inducement of a policy change, as opposed to a "counter" that only offsets the impact of the measure. With this distinction in mind, let's look at the issue of subsidies and dumping separately.
Starting with subsidies, these are generally considered to fall into the category of non-market policies and practices, although perhaps only certain kinds or amounts of subsidies would qualify. "Countervailing duties," kind of by definition, are a tool to counter those subsidies. But the question I have is, can they ever induce a change in the foreign government's subsidy policy or practices? My sense is that policy change almost never happens as a result of CVDs (but if there is any empirical research on this, or people have good anecdotes, I'd love to hear about it). Thus, if the goal is to induce the subsidizing government to stop the subsidies, which is what at least some people are hoping for I think, CVDs don't seem to have much impact. (One possible counterexample I can think of here is Canadian lumber, where I think the Canadians have tweaked their system a bit over the years, but even there the underlying dispute has not been resolved and the general program has continued.) On the other hand, if the goal is simply to "level the playing field," then CVDs can achieve this by imposing a tariff to restrict imports.
As to dumping, there is an argument that dumping is caused by domestic protectionism or various other non-market policies and practices. However, identifying such policies and practices is not generally part of an anti-dumping investigation, so it's hard to see how anti-dumping could address these policies and practices effectively. There are some special provisions in the anti-dumping law of the EU and U.S. (and perhaps others) that target certain non-market economy behavior, and can lead to higher anti-dumping duties where it exists, but again my question would be, has this ever led to changes in the non-market policies and practices? I've never heard of that happening.
From what I can see, if the goal is to induce changes to non-market policies and practices, the best option, which has worked many times over the years, is filing WTO/FTA complaints directly against those policies and practices. A neutral international ruling seems to have more impact here than unilateral remedies such as AD/CVDs, although of course some of these policies and practices are just too sensitive in terms of domestic politics to change even with an international ruling.