This interesting paper touches upon a wide range of issues. The authors remind us of the importance of psychological factors in Keynes' view of how an economy works; propose an application of state-of-the-art stochastic techniques to issues of government deficits and debt repudiation; and argue that nonlinear policy reactions can, even in a rational expectations framework, produce results that are somewhat reminiscent of Keynesian phenomena.

The three short sections below deal with the first point; with the general features of the technique proposed; and with their applicability to fiscal policy and debt management.

Psychology vs. rationality

Forward-looking behaviour has an essential role in correctly specified macroeconomic models. Investment depends on expected future profitability, asset prices on expected future dividends and capital gains, consumption on expected future incomes, and so on.

In Keynes's view, the future can be so unmeasurably uncertain as to make it impossible to specify objective probabilities for the relevant realizations of future variables. Expectational variables are then essentially subjective, and can shift in arbitrary ways as agents change their mind. In this framework, any phenomenon could be interpreted (but not really explained) in terms of exogenous, unpredictable expectational shocks – an unpleasant state of affairs from an economist's point of view. The rational expectations school set out to pin down expectational variables: taking the probability distribution of future exogenous variables and the structure of the economy as given, it becomes possible for agents to compute objective probability distributions and expectations, and for economists to undertake prediction and normative analysis.