Topics in macroeconomics – non-standard preferences and macroeconomics

Most macroeconomic models build on time separable preferences with geometric subjective discounting, i.e., individuals maximize

Ample empirical evidence shows that this is not consistent with how individuals actually make their choices. In this course, we will study a number of deviations from standard preferences and how these affect macroeconomic outcomes. When this is written, the plan of the course is not fully determined, tentatively; we will cover the following examples of non-standard preferences and their implications.

• Epstein–Zin preferences which allows a de-coupling of the intertemporal elasticity of substitution and risk aversion, while these conceptually different entities are determined by the same parameter under standard preferences. We will study to what extent these preferences can help understand important puzzles in asset pricing.
• Time-inconsistent preferences. Evidence from experiments and the psychological literature suggests that the assumption of geometric discounting is inconsistent with actual behavior. A classical example is that of preference reversal. Individuals are asked at time 0 to choose between two goods x and y delivered at, say time s and s+1, respectively. These goods are chosen so that individuals have a slight preference for y, delivered on period later. Very often it happens that as time goes preference reversals occur. For example, at time s, individuals who at time 0 preferred good y, now prefer x today rather than y tomorrow. To account for such reversals, David Laibson has brought the idea of quasi-geometric (or quasi-hyperbolic) discounting into macroeconomics. Robert Barro and others have shown that the existence of such preference have important normative and positive implications for macroeconomics.
• Evidence shows that individuals attach substantial weight to reference points and seem to perceive utility losses if income or consumption falls below the reference points that are much larger than the utility gain if it is above. These effects are qualitatively and not the least quantitatively very different from standard risk aversion. They can potentially help explain behavior that seems inconsistent with reasonable degree of risk aversion and also of policy inertia, i.e., the difficulty of reversing a political decision on, say transfers to groups of individuals.
• From introspection it appears very unreasonable that the utility of consuming a particular good today is independent of previous consumption. Habit formation, addiction and the build up of tastes for particular goods seem to be important for current consumption choices. We will study examples of how such preferences have been modeled and how they have been used to explain empirical observation.
• Gul and Pesendorfer have recently developed a theory of preferences that allows individuals to have preferences over the options they face. They show that under such preference concepts like resistance or giving in to temptations can meaningfully formalized.