While research has shown how current inflation can shape people’s expectations of future inflation, we know little about whether or how such expectations are transmitted to future generations. Understanding inflationary expectations across generations is important for central bankers and for researchers hoping to properly account for this phenomenon in their policymaking and modeling.
The authors study this question by focusing on Germany, a country that experienced severe hyperinflation in 1922-23 and which still resonates today, at least anecdotally. To bring some empirics to bear, the authors collect data on local inflation of 633 German towns between 1920 and 1924, as well as data on contemporary inflation expectations of households from three large-scale surveys. They link today’s inflation expectations with historical inflation either at the zip code level or the county level. The merged data allow for comparisons of households whose ancestors likely experienced different inflation in the early 1920s but have the same observable characteristics today, are surveyed at the same point in time, and live in the same region. Before describing the finding from this analysis, let us first consider suggestive evidence from across Europe.
Figure 1 shows households’ inflation expectations in 2022 for 2023 (Panel A), as well as realized inflation in 2022 (Panel B) separately for European countries that did not experience a hyperinflation before 1930 (left bars) and for European countries that experienced a hyperinflation before 1930 (right bars). Panel A reveals that individuals living in countries with a hyperinflation in the distant past have inflation expectations that are approximately 1.4 percentage points higher than inflation expectations of individuals living in countries without such a history. In contrast, Panel B shows the difference in realized inflation rates between the two groups of countries to be only 0.3 percentage points and thus very similar. Thus, we can see that inflation expectations are shaped by inflation shocks in the distant past rather than only by current inflation.
As for Germany, Figure 2 illustrates the authors’ main finding. The authors sort German towns into quintiles based on their cumulative inflation between 1920 and 1924. For each quintile, they compute the average inflation expected by households living in those towns today, they then plot average inflation expectations against historical inflation quintiles. Consistent with Figure 1, the authors find a strongly positive relationship between historical inflation and contemporary inflation expectations, suggesting that individuals living in towns with high inflation in the 1920s expect higher inflation today. Moving from the quintile with the lowest inflation to the quintile with the highest inflation increases expected inflation by 0.7 percentage points, which corresponds to around 90% of the average realized annual rate of inflation during the time period covered by the surveys. This result points towards a long-lasting effect of inflationary shocks on inflation expectations. (Polish households residing in formerly German areas reveal similar effects.)
The authors consider two likely mechanisms for how German hyperinflation persists locally across generations:
Bottom Line: Inflation expectations appear transmittable across generations. Importantly for policymakers today, this finding suggests that the recent surge in inflation will have lasting effects, resulting in elevated inflation expectations around the globe for a prolonged period of time, even after current inflation rates recede.