Steinar Holden, ECON 4325
Key elements of New Keynesian models:
(based partly on Woodford (2009), Chari, Kehoe,
McGrattan (2009) and Gali chapter 8)
 Expectations are
o important (for consumption, investment,
employment, wage and price setting) and
o should be endogenous
 model consistent (rational) or perhaps
learning, or predictable irrationality
o communication important part of monetary
policy
 Use models with coherent intertemporal general
equilibrium foundation
o consistency between short run and long
run issues
 Real disturbances are an important source of
economic fluctuations
o technology (aggregate productivity and
investment specific) preferences,
government policies, etc.
o Implies that some fluctuations are efficient
(fluctuations in natural levels of output and
interest rate)
1
o Natural level of output should not be
modelled as trend output, nor should
output gap be modelled as deviation from
trend
o Not efficient to stabilize the economy
completely
 But there are also inefficient fluctuations
o Associated with sticky prices and wages,
(but presumably also other mechanisms)
 Monetary policy can be used to stabilize the
economy – in particular inflation, but also
dampen fluctuations in output
o Control of money stock is not necessary as
instrument
 Desirable to use econometrically validated
structural model
o More eclectic approach to estimation of
model parameters
o Fairly strong priors from theoretical
consistency
2
Some remaining issues
 Wage and price stickiness is taken as exogenous,
given by Calvo framework (time-dependent price
setting)
o Problem – if long time since adjustment, price
or wage will be far from optimal value,
involving a large efficiency loss due to large
effect on output/employment
o Is this plausible?
 If firms may change price or wage subject to small
cost (menu costs; state-dependent pricing), many of
the key features of the model remain, but the
welfare implications are changed considerably
o Firms with large deviation from optimal price
will be more likely to change price, which
reduces the costs associated with price
dispersion and thus also the costs associated
with inflation
o If menu costs are small, the costs associated
with distortions of relative prices will also be
small.
3
 An aside – are wage and price stickiness only
inefficient distortions?
o Wage/price rigidity to ensure efficient
investments by avoiding holdup problems,
where one party exploits the ex post bargaining
position to capture share of return from other
party’s investment
o Wage/price rigidity as insurance, e.g. to reduce
uncertainty for risk averse workers
 Does labour demand depend on contemporaneous
wage?
o Alternative: implicit contract where
o employment decisions are intertemporal,
depend on present value of marginal revenue
relative to present value of wage costs, and
where the wage level is smoothed to insure risk
averse workers
o separations are still efficient (i.e. take place
when outside option of worker exceeds
marginal revenue of firm)
4
 As New Keynesian theory has aimed for closer
correspondence with data, theoretically less
appealing features have been introduced
o Backward indexation of prices to allow for
non-zero steady state inflation (is not
consistent with evidence on price setting)
o Introduction of new shocks, like wage markup
shock, price markup shock, exogenous
spending shock, risk premium shock, etc
 But aggregate productivity shock is also
problematic
o Taylor-rule is not consistent with past
monetary policy setting, because a large
random component need be added (Chari et al)
 Importance of the interest rate
o Interest rate expectations affect important
irreversible decisions like real investments, e.g.
the purchase of a house/building of a factory
o Expectational errors may be extremely costly
o Suggests that central bank interest rate should
be predictable, and also try to avoid large
deviations from “normal” levels
 Financial stability important
o Asset variables may deviate from equilibrium
levels, involving large imbalances, leading to
strong volatility in output and employment
o Credit crunches and liquidity squeezes may
have strong negative effects on output
5
 Interaction with fiscal policy
o Expansionary fiscal policy leads to increased
output and higher inflation, which is
counteracted by a rise in the interest rate =>
dampens expansionary effect
o Correspondingly, monetary response dampens
the negative effect of fiscal contractions
 Zero lower bound for the interest rate
o Puts a lower bound for the real interest rate
when recession is combined with low rate of
inflation
o If possible, avoid fiscal contractions as long as
the monetary policy is constrained by the zero
lower bound
 New Keynesian framework usually does not allow
for asymmetric fluctuations, where downturns are
more persistent or larger than upturns
o If economy does not work due to severe
frictions or distortions, excess downturns may
result
o Implies that business fluctuations may have a
negative effect of average output over a longer
period, in contrast to standard assumptions
o May strongly amplify the gains from vigorous
economic policy to counter downturns.
6
 In economies with large wage setters, there will be
strategic interaction between wage setting and
monetary policy
o Strict inflation target may lead to lower
equilibrium rate of unemployment
 Wage setters know that high wage growth
will make CB set high interest rate
 Wage setters moderate their wage claims
to avoid interest rate hike
 Reduced wage pressure leads to lower
equilibrium rate of unemployment
o But wage setters’ incentives to coordinate
wage moderation may be greater under a more
passive monetary regime, e.g. a monetary
union
7