Essays on Durable Goods and Housing in New Keynesian Models
Abstract
While they are widely used for policy, two sector New Keynesian
models with flexibly priced durable goods are rejected by the
data. After a monetary policy shock,these models typically
demonstrate negative co-movement between durable and nondurable
consumption. This is completely at odds with VAR evidence which
suggests that they co-move positively (and that durable good
consumption is substantially more sensitive than non-durable good
consumption to the shock). This thesis focuses on and explores a
number of possible solutions to this puzzle and ultimately
proposes a model that resolves the puzzle and retains many of the
appealing attributes of modern macroeconomic models for policy
analysis.
After an introduction that describes the durable good co-movement
puzzle in greater detail, chapter 2 examines the potential of a
simple production linkage - where nondurable goods are used for
consumption but also as inputs into the production of durable
goods - to resolve it. This approach is simple and intuitive.
But, while it generates positive co-movement, it struggles to
capture the responsiveness of durable goods to the shock.
Chapter 3 examines whether restrictive assumptions on the demand
side might be hindering the model’s capacity to be accepted by
the data. Specifically, it uses the Kimball (1995) aggregator to
introduce smoothed kinked demand curves for both types of goods.
This yields an augmented New Keynesian Phillips’ curve that
dampens the response of prices to changes in marginal cost and,
as a result, a model capable of generating positive co-movement
at levels of nominal price rigidity far lower than in standard
New Keynesian models.
Chapter 4 examines how the introduction of a borrowing constraint
into the standard two sector New Keynesian model affects the
generation of positive co-movement. While earlier works have
shown that this extension actually makes it more difficult to
generate positive co-movement, they have been far less conclusive
in diagnosing why. According to one theory, when the constraint
tightens after an unanticipated
increase in the policy rate, savers find that they are prevented
from saving as much as they would like and, as a result, are
forced to transfer consumption into the future through the
purchase of durable goods (thus exacerbating the negative
co-movement). The model in this chapter considers this contention
and demonstrates that, even when they are given other ways to
save, savers prefer to purchase durable goods. The desire to
purchase of durable goods is so strong that they even borrow in
these alternative markets in order to do so.
Chapter 5 uses the introduction of nominal wage frictions to
reconcile the model with the borrowing constraint with the
empirical literature. It also extends the model to include a
fully specified financial sector and macroprudential policy. (It
also examines some of the policy trade-offs inherent in
macroprudential policy and monetary policy that targets financial
aggregates.) In this sense, the model in this chapter offers a
coherent and consistent workhorse model for future policy
analysis.