With the help of empirical methods, laboratory experiments and theoretical models, the junior research group aims to improve our understanding of macroeconomic expectations and to investigate the implications of expectation formation for monetary and fiscal policy. There will be a particular focus on the analysis of new data on expectations collected during the times of uncertainty and economic crises caused by the COVID-19 pandemic.
The current economic situation is unique, since, unlike in previous economic crises, there is a clear external non-economic cause that is driving economic downturns. Especially valuable insights can be gained from analyzing how expectations respond to news. Of particular interest will be news about measures that may mitigate the consequences of the COVID-19 pandemic, such as vaccines and successful social distancing. In addition, virtually all countries are affected by the virus, making this crisis the most global economic crisis of all time. Finally, different countries and regions are affected at different times, leading to exogenous variation in the data that is invaluable for econometric analysis.
Survey data on expectations are currently leading to new insights and hypotheses about the formation of expectations, to which the junior research group will contribute. These hypotheses can be futher tested and refined in an artificial, controlled environment by means of laboratory experiments with human subjects. Any new insights from the above two metodologies can then be used in macroeconomic models. This will help to get more robust answers to questions such as: ‘How can monetary and fiscal policy speed-up the recovery from a major economic crisis?’, ‘How can the central bank anchor expectations to its target and keep inflation stable?’, ‘How can stable and low unemployment be achieved?’, and ‘How should monetary and fiscal policy makers deal with heterogeneity in expectations?’
State of the literature
Expectations from households, firms and other economic actors play a crucial role for developments in the macro economy. However, empirically, the assumption of fully rational expectations used in most macroeconomic models does not sufficiently correspond to the actual expectations of households and firms.
Several alternative modeling approaches assume that at least a part of expectations are backward-looking. However, the literature on the empirical validation of backward-looking expectation formation is small and not entirely convincing. Backward-looking expectations create persistence in expectations and are, therefore, a possible explanation for the greater persistence found in survey data compared to rational expectations. However, there is no convincing evidence that backward-looking components are the correct explanation for the higher persistence of survey expectations and that expectations are actually formed based on past observations.
Instead, recent empirical studies on survey expectations tend to suggest that expectations in survey data are ‘sticky’. This means that forecasters base their new forecasts for a given future period in part on forecasts for the same period that they made a quarter earlier. Forecasters are, therefore, revising their expectations in a dampened manner.
Another bias in survey data that makes expectations more persistent could be that individual forecasters partially base their forecasts on the last observed median forecast of other forecasters. Other (lagged) sources of information on which individual forecasters may base their own forecasts include projections published by policy makers such as central banks.
Although the above biases make expectations more persistent, at the same time, it is found that people overreact to new information. For example, with an extrapolation bias where forecasters excessively extrapolate the latest economic news into the future.
Another important aspect of macroeconomic expectations is heterogeneity. Substantial differences between the expectations of households, firms and professional forecasters are found in survey data. And there are also large differences between different households and between different firms.
Methodologies and findings
The above new findings are part of a burgeoning literature on the empirical analysis of expectations from survey data and laboratory experiments. The considerable advances that have been made in this literature in recent years have opened the door for new research endeavors, and the junior research group should contribute to this. To this end, the junior research group relies on three methodological pillars.
The first pillar is the empirical analysis of expectation data from different surveys. The focus of this pillar will be on the analysis of very recent data which is currently still being produced. Valuable new insights into the formation of expectations can be gained by analyzing how macroeconomic expectations are reacting to the COVID-19 pandemic.
The second methodological pillar of the junior research group will be controlled laboratory experiments. In such experiments, a researcher can exogenously vary a certain part of subjects’ economic environment or information set across treatments, while everything else remains constant. This methodology is, therefore, ideal for the further development and testing of theories and hypotheses about expectation formation that were obtained from the empirical analysis of survey data.
The third methodological pillar is about incorporating new findings about expectation formation from survey data and laboratory experiments into macroeconomic models. This building block is fundamental, because macroeconomic models can be used to make statements about, a.o., how central banks should set their interest rates and what kind of fiscal policy is most effective. This pillar therefore connects the first two pillars to monetary and fiscal policy implications.
Together, the three methodological pillars can help us to eventually arrive at a unified theory of expectation formation that is compatible with empirical findings and can explain actual macroeconomic expectations. With such a more realistic unified theory of expectation formation,
great progress can be made when it comes to providing effective and robust monetary and
fiscal policy recommendations. This will help policy makers to better respond to economic crises such as the 2007/2008 financial crisis and the current crises triggered by the COVID-19 pandemic. More generally, the theory may help to achieve a more stable economic environment with low unemployment and inflation.