That conclusion was reached largely because, if I use Professor Goldman's terms, it's difficult, generally speaking, to change the cost-benefit calculations of certain kinds of governance.
Among the interesting cases of sanctions success historically, two cases are usually held up. One involves the sanctions imposed against the State of Israel by the Eisenhower administration in 1956. It was a threat of sanctions rather than the actual imposition of sanctions. The Eisenhower administration threatened to revoke the State of Israel's status under U.S. income tax law, as a tax-deductible donation, as a mechanism for forcing Israel to shift its policies on the Sinai. The Israeli government moved immediately.
The second case of a liberal democratic government sanctioning another liberal democratic government was in the aftermath of the French bombing of the Greenpeace ship in Auckland harbour in the mid-1980s. The New Zealand government seized the French agents. France let it be known to New Zealand that if these agents were not released into French custody, New Zealand exports to the EU would be negatively affected. New Zealand understood entirely what that threat involved. The agents who had planned and executed that bombing were released into French custody, released by French authorities, and indeed given medals for their behaviour.
There aren't that many cases where you get those kinds of sanction episodes, but in the case of liberal democratic governments, it seems to me that you do find, especially if there is dependence, a sensitivity there that just simply isn't the case in authoritarian or military dictatorships.