Experimental Design
We run a survey among German households and examine whether changes in monetary policy expectations shift households’ spending intentions.
Our survey elicits prior expectations on the 2023 inflation rate and on the future kind of monetary policy employed the ECB.
For causal interpretation, we rely on three disjoint treatment arms that aim to exogenously shift households’ expectations about monetary policy. The first treatment informs participants that the ECB has raised the interest rate (without any information on the level of rate of increase). Participants in the second treatment arm receive the information that we ran a survey among economic experts, and that the interest rate increase by the ECB decreased the short-term inflation expectations of those experts by 1.2 percentage points. The participants in the third treatment arm receive the information that we ran a survey among economic experts and that those experts told us that an increase in the interest rate by the ECB decreases the inflation rate, because the demand for goods decreases. We employ an active control group design. Participants in the control group receive (irrelevant) information about the population development in Germany.
We then asked participants about their posterior expectations on the inflation rate and on the future kind of monetary policy employed the ECB. We also asked participants question on spending intentions on durable and non-durable goods as well as whether now is a good timing of these purchases.
Our main hypothesis is that the treatments in a first stage impact participants’ expectation about future implementation of monetary policy. We believe that they will find it more likely that monetary policy will be restrictive. The second hypothesis is that the treatments in the second and third treatment arm give rise to higher treatment effects comparted to the treatment in the first treatment arm.
We then assess in 2SLS-design the causal effect of the change in monetary policy expectations on spending intentions. Our third hypothesis is that treated participants reduce their spending intentions.