Recent research has connected fiscal policy during the pandemic with subsequent inflation. One common piece of evidence is the close relationship between “excess savings” and inflation in recent years. Some of our friends around the Fed discuss this in more detail: Fernando Martin, Hamza Abdelrahman and Luiz E. Oliveira, and Omar Barbiero and Dhiren Patki.

These studies calculate the level of excess savings by adding up all the savings during the pandemic that deviate from the pre-pandemic trend. Here, we use an alternative measure of household savings that also considers changes in asset valuation and leverage: specifically, the difference between household total assets and total liabilities, known as household net worth. FRED gets the relevant data from the US financial accounts published by the Board of Governors of the Federal Reserve System.

One advantage of using household net worth is that it considers the value of housing wealth, or home equity. This component of household wealth is crucial—and maybe even more so in this period because of the significant increase in house prices, which implies a considerable increase in housing wealth.

The FRED graph above shows the evolution of household net worth (divided by the level of disposable personal income to represent its relative size) and the annualized quarter-to-quarter PCE inflation rate.

Our graph confirms the close connection between household savings and inflation during this period and shows that, in the third quarter of 2023, household wealth remained above pre-pandemic levels.

How this graph was created: Search FRED for and select the “household net worth” series. From the “Edit Graph” section, use the “Edit Line 1” tab to search for and add the “disposable personal income” series. Apply formula 100*(a/(b*1000)) and click “Apply.” For the second line, open the “Add Line” tab, search for and select “PCE Index” and apply formula 100*((1+a/100)^4-1). (For consistency, we multiply disposable personal income by 1000, we annualize the quarter-to-quarter percent change in PCE, and we multiply both indicators by 100 to display them in percent.) Finally, use the “Format” tab to move the y-axis for the first line to the right side and start the sample period in 2016.

Suggested by Masataka Mori and Juan M. Sánchez.