Studies that promote investor-state dispute settlement (ISDS) have generated misleading numbers and dubious recommendations. Legislators should approach all studies with caution, warns Gus Van Harten.
Gus Van Harten is associate professor at Osgoode Hall Law School in Toronto, Canada, and an expert on ISDS. His research is freely available here.
If they looked closely at investor-state dispute settlement (ISDS) in TTIP and the Canada-EU CETA, I expect that most voters and taxpayers would be amazed – and not in a good way.
ISDS transfers fundamental powers from elected legislatures and independent courts to private lawyers, sitting as arbitrators. Essentially, it requires the public to subsidise foreign companies for risks of democratic regulation, in ways that no country’s constitution and courts would require. That makes ISDS a tough sell.
Personally, I don’t think ISDS, based on the European Commission’s model, has any place in TTIP or CETA. More importantly, the public should know that proponents of ISDS – including public officials – have often used inaccurate information to downplay its impact on legislatures and regulation.
A good example is a report commissioned by the Netherlands trade and foreign affairs ministries and submitted to the Dutch parliament in 2014. The authors of this report relied on false or misleading information to downplay concerns about ISDS and regulatory chill. For instance, they said (I have underlined the clearest inaccuracies):
The second argument against regulatory chill is that most ISDS claims do not challenge legislative acts[…] In a study published in April 2014, researchers Jeremy Caddel and Nathan Jensen concluded that the vast majority of investor-state claims arise from executive branch decisions instead of legislative decisions.
After analysing all concluded ISDS decisions, the researchers found that 47% of disputes were associated with ministries or agencies while only 9% (14 total cases) resulted from legislative acts.
These statements were based on a study by two US political scientists (Caddel and Jensen) that was published shortly before the Dutch report. In the study, Caddel and Jensen said that “there were only 14 cases of legislatures taking actions leading to disputes” in 163 ISDS arbitrations at the World Bank’s International Centre for Settlement of Investment Disputes. In turn, they made a dubious recommendation that was flagged in the Dutch report:
Given the low rate of disputes involving legislative branch activity, arguments that investor-state arbitration may encroach on the legitimate prerogatives of domestic governments appear to be overstated. Instead, democratic legislatures should embrace investor-state arbitration as an additional check on executive branch misbehavior.
The trouble is, the numbers were wrong. There are far more than 14 cases of “legislatures taking actions leading to ISDS disputes”. And they account for much more than 9% of overall cases.
I say this with confidence because, in a study of 162 investment treaty cases (up to June 2010) I found 60 cases in which a legislative decision was challenged by a foreign investor. Many of these cases were brought against countries with well-established democratic institutions.
My findings and data were published about six months before the Caddel and Jensen study and the Dutch report. They were not discussed in either.
How then could our numbers – 60 versus 14 – be so different? The answer lies in the methods used to count cases.
Caddel and Jensen classified each case based on their estimation of the “primary” decision that was challenged by the foreign investor. As a result, their numbers did not include many cases in which a foreign investor targeted a legislative decision alongside an executive decision.
On the other hand, I classified each case based on all of the decisions that were challenged by the foreign investor. If any one decision was a legislative decision, I counted it as such. I did not try to vet any one decision as “secondary” to another that was challenged, partly because this involved more researcher discretion and uncertainty.
This led to my finding that about 37% of the cases involved a legislative decision compared to Caddel and Jensen’s of 9%. That’s a huge disparity, arising simply from a difference in the methods.
I have no quarrel with Caddel and Jensen’s choice of methods and research priorities. My complaint is that they did not mention the effect of their choice on the numbers, when they presented the findings publicly and made their bold recommendation to legislators.
Any study – including mine of course – should be read with care to evaluate the methods and assess the relevance of its findings and the credibility of its conclusions. That is especially true for public officials and publicly-funded researchers, such as the authors of the Dutch report.
In this case, the worst thing is Caddel and Jensen’s misleading numbers were provided to legislators on the essential question of how often ISDS has been used to challenge legislative decisions. Also, this was not the only significant inaccuracy in the Dutch report. Frankly, it would take many articles like this one to explain them all.
I do not wish to leave the impression that all legislative decisions are challenged or chilled by ISDS. But the prospect of this happening under TTIP and CETA is real, and it carries important risks for voters and taxpayers. The issue looms largest when proposed laws are resisted by major companies, creating an uncertain risk of potentially massive and unavoidable costs for the public.
Finally, the onus is on the proponents of ISDS to make credible claims. They are the ones trying to justify the profound transfer of power and selective subsidy that are inherent in ISDS. It’s a bad sign when proponents use inaccurate information to do that.