Very happy that our paper “The Transmission of Monetary Policy under the Microscope” has been published in @JPolEcon, joint work with @BlomhoffHolm and @ATischbi. With a slight delay, here is a short summary! https://t.co/sUDUHuPM5V (1/N)
4
45
265
Replies
What’s the effect of monetary policy on people’s consumption, income, saving? Recent HANK models provide valuable theoretical insights, but empirical evidence remained scarce. We therefore use detailed administrative data from Norway to address this question. (2/N)
1
0
1
A number of HANK models make the following predictions: (1) Households with low liquid wealth should have high MPCs and large consumption responses to income changes, (2) MP influences HH consumption largely through indirect, general equilibrium effects affecting HH income (3/N)
1
0
1
In a nutshell, we confirm these predictions, but also find evidence that is inconsistent with typical HANK models, providing new challenges! Here is what we find. Whether we sort HHs by liquid assets or their net interest rate exposure, we get a similar picture. (4/N)
1
1
2
Following a monetary policy tightening (higher interest rates), people at the bottom of the distribution (low liquid assets) see their income fall which strongly feeds into their consumption and this response strengthens over time (Prediction #1:✅). (5/N)
1
0
1
However, what’s surprising is the response at the top of the distribution (high liquidity HHs). Their consumption rises when interest rates increase! What’s going on? These people have vast amounts of liquid assets that pay higher interest rates after a MP tightening. (6/N)
1
0
2
The financial income of these people therefore increases, and with relatively high MPCs, they consume the additional income. Challenge #1: Typical HANK models feature low MPCs for high-liquidity HHs and would predict only small consumption responses. (7/N)
1
0
2
Next HANK prediction: monetary policy affects consumption mainly through indirect general equilibrium effects. Using an empirical decomposition into direct and indirect effects, we find that indirect effects dominate (Prediction #2:✅). (8/N)
1
1
3
However, initially, direct effects outweigh and indirect only kick in after some time when wages start to move. In contrast, most HANK models would feature an immediate dominance of indirect effects➡️ Challenge #2. (9/N)
1
0
4
We wrote a longer VoxEU summary some time ago: https://t.co/4fl7a9YsaQ Also: there is lots more in the paper, check it out! We’ve also made all our codes & MP shocks publicly available: https://t.co/8jBas9GQk6 (10/N)
1
1
10
Last: Norway not only has great data, but also bright and sunny summers! Worth a trip! (N/N)
1
0
13
@pascalpaul @JPolEcon @BlomhoffHolm @ATischbi Hi Pascal, congratulations to you and the coauthors, this is quite an achievement.
0
0
2
0
0
0