Thomas Dietz, Adam Smith, rationality and manipulation by unscrupulous agents

In Nature (In retrospect: The art of influence), Thomas Dietz "reassesses" Robert Cialdini's 1984 book Influence, which argued (again) that decision-making is often irrational and subject to "manipulation by unscrupulous agents."

We make decisions based on narrow self-interest, calculated benefits, costs and risks. Or so claimed economist Adam Smith, whose 'rational actor model', from his 1776 opus The Wealth of Nations, has long dominated thinking in economics and social science. By the late twentieth century, Smith's view had been applied to every domain of human decision-making, from marriage to international negotiations. But a growing body of evidence began to indicate that the model was often misleading.

Adam Smith claimed that we sometimes make rational, self-interested decisions (as exemplified by an unscrupulous salesman trying to manipulate his customers' wishes) and that a model that we always decide that way would often be misleading (as exemplified by those customers who, against their own interests, fall victim to the salesman's tactics). He said it in 1776. In 2011, Dietz agrees with him. But he writes as if Adam Smith never said such things.

Cialdini saw that Smith's model of rational decision-makers, immune to any influence other than information, was simplistic.

Mythical Adam Smith is, to say it in improper English, one of the strawest men ever.

Mainstream policy analysis still relies heavily on the assumption of a rational decision-maker, but social psychology is starting to affect how policies are designed. In the 2008 book Nudge (very much a descendant of Influence), authors Richard Thaler and Cass Sunstein argue that insights from the social sciences — such as our strong tendency to choose the default option no matter what it is — can be used to encourage better decisions.

This paragraph doesn't make sense. Thaler and Sunstein's proposal would replace the current system of government coercive paternalism with one they call "libertarian paternalism" based on free, but subliminally directed, choice. In either case paternalism assumes that people are irrational and thus unable to make wise choices by themselves. In what sense does "mainstream policy analysis still rely heavily on the assumption of a rational decision-maker"?

Dietz later makes the good point that although "libertarian paternalism" may be better that outright coercion, "unscrupulous agents" in government may use the salesman tactics Cialdini documents in his book to "manipulate us against our own interests."