Asymmetric Monetary Policy Tradeoffs

A woman and a man try to hold down a gigantic balloon that is trying to float away

The presence of an inflation-unemployment tradeoff is at the heart of monetary policymaking. More recently, questions regarding this tradeoff became particularly relevant as inflation in the US and Europe surged to its highest level since the 1970s. 

In the BSE Working Paper 1404, “Asymmetric Monetary Policy Tradeoffs,” Davide Debortoli, Mario Forni, Luca Gambetti, and Luca Sala use a novel nonlinear Proxy-SVAR approach to measure the inflation-unemployment tradeoff associated with monetary easing and tightening during booms and recessions for the US economy. 



Non-linearities play a critical role in shaping the overall impulse responses to a monetary shock

Figure 1 plots the impulse response functions (solid lines) to a monetary shock for the nonlinear model. For all variables, the responses of the nonlinear components (second and third columns) are similar in magnitude to those of the linear counterparts and, in most cases, significant. In particular, the nonlinear components have persistent and significant positive effects on unemployment (fourth row). This implies that monetary policy leads to larger changes in unemployment if a tightening is implemented during a recession. In contrast, for prices (fifth row), the nonlinear components operate in opposite directions, implying that the largest changes in inflation are associated with monetary policy tightening.

Notes: Impulse response to a monetary shock of the linear (first column), sign (second column) and state (third column) components. Solid lines represent point estimates, the gray areas are 68% confidence bands.
Figure 1. Impulse Response Functions in the Nonlinear Model

Notes: Impulse response to a monetary shock of the linear (first column), sign (second column) and state (third column) components. Solid lines represent point estimates, the gray areas are 68% confidence bands.

Monetary tightening has large and significant effects on real variables in booms and recessions.

Figure 2 compares the total effects of monetary easing and tightening during booms (columns 1 and 2) and recessions (columns 3 and 4). For real variables (unemployment and industrial production), monetary tightening always has large and significant effects. Instead, monetary easing has much smaller effects. The authors argue that this result could imply that the central bank should adopt more aggressive measures to fight a recession. They also note that the desirability of a more aggressive stance crucially depends on the inflation-unemployment tradeoff facing the central bank.

Impulse response to a monetary shock of the linear (first column), sign (second column) and state (third column) components. Solid lines represent point estimates, the gray areas are 68% confidence bands.
Figure 2. Impulse response functions in the nonlinear model

Notes: Impulse response to a monetary shock of the linear (first column), sign (second column) and state (third column) components. Solid lines represent point estimates, the gray areas are 68% confidence bands.

Countercyclical monetary policy is associated with relatively favourable inflation-unemployment tradeoffs

Figure 3 displays a scatterplot of the (cumulative) impulse responses of inflation and unemployment, where each point corresponds to the cumulative effect over alternative horizons. This figure illustrates three interesting results. Firstly, monetary easing (red solid line) during recessions leads to a protracted and significant decline in the unemployment rate and to no significant change in inflation. The authors highlight that this result shows that monetary easing could be an effective tool to stimulate the economy in recessionary periods. Secondly, deflationary policies during expansions (blue dashed line) have moderate costs in terms of unemployment. However, these costs are substantially smaller than those associated with contractionary policies during recessions (red dashed line) or those implied by a flat Phillips curve (which would be infinite). Lastly, monetary easing in an expansion is very inflationary with virtually no effects on the unemployment rate – i.e., an extremely large inflation-unemployment tradeoff.

The figure plots the relationship between the cumulative change in inflation (x-axis) and in the unemployment rate (y-axis) at different horizons (H={12,24,36}), in response to monetary easing (solid lines) and tightening (dashed lines), during booms (blue lines) and recessions (red lines).
Figure 3. The Effects of Monetary Shocks on Inflation and Unemployment

The figure plots the relationship between the cumulative change in inflation (x-axis) and in the unemployment rate (y-axis) at different horizons (H={12,24,36}), in response to monetary easing (solid lines) and tightening (dashed lines), during booms (blue lines) and recessions (red lines).

Overall, these results provide support for the use of countercyclical monetary policies as they are associated with relatively favourable inflation-unemployment tradeoffs. In addition, the authors also show that these results can be rationalised by a simple model with downward nominal wage rigidity, which is used to assess the validity of their empirical approach.