Federal Reserve Economic Data

The FRED® Blog

Counting the wealthiest 0.1% of households

Recounting how the DFA data are reported

The FRED Blog uses graphs to help tell the story behind the data. But sometimes the FRED graphs themselves tell us there’s even more to the data than meets the eye. Today, we look at wealth data based on households interviewed in the Survey of Consumer Finances (SCF) and reported in the Distributional Financial Accounts (DFAs) from the Board of Governors of the Federal Reserve System.

The solid blue line in the FRED graph above shows the number of U.S. households in the wealthiest 0.1% of the population, with units on the left axis. The dashed red line shows the sum total of all the other U.S. households, with units on the right axis. Although the household count of the wealthiest households has increased between 1989 and 2022, it hasn’t done so steadily or continuously. And that caught our eye.

The number of households in the wealthiest 0.1% of the population is a proportion of the total number of households interviewed in the SCF and reported in the DFAs. So, when the total number of households increases, so does the number of households with the most wealth. That certainly seems to be the case between 1993 and 2020.

Of course, the same applies in reverse. Consider the first three quarters of 2020—the early stages of the COVID-19 recession. During that time, due to socioeconomic reasons, the overall number of households decreased. And so did the count of households in each of the five wealth groupings reported in the DFAs. That included the number of the wealthiest, the top 0.1%.

However, there was no such decrease in the overall number of households between 1989 and 1993, when the count of wealthiest households fell by 8.6%. So how can this be explained? The short answer is that it’s an artifact of using a survey (the SCF) to estimate the characteristics of a small group (the top 0.1% in the DFAs).

Not satisfied by the short answer and want to learn more? Read on!
The longer answer is that the overall sample size of the SCF was smaller in earlier years (the 1990s) than in later years (the 2000s). More importantly, the number of the wealthiest, the top 0.1%, interviewed in the survey was relatively smaller than it is in more recent surveys. This means that the DFAs report larger proportions of the wealthiest households in those earlier surveys.

When determining where the threshold lies for the top 0.1% of wealthiest households, the DFAs create the group by assigning a finite number of households from the SCF to the top 0.1%. The threshold will get as close to the top 0.1% as the data allow, but it will not necessarily match an individual household in the population. In that way, in 1989, the SCF threshold lies at 0.11% (too many households); and, in 1992, it lies at 0.09995% (too few households). That is why the DFAs show a declining number of households in the top 0.1% during the first several years of the survey.

Another way to put it is that small differences in proportions, when applied to populations of almost 100 million households, make for eye-catching patterns in reported numbers.

How this graph was created: Search FRED for “Household Count in the Top 0.1% (99.9th to 100th Wealth Percentiles).” On the resulting graph, click on “Edit Graph,” open the “Add Line” tab, and search for the same series again. For this line, then search for and add the other series, then apply formula a+b+c+d. Finally, in the “Format” tab, place the legend for the second line on the right side.

Suggested by Diego Mendez-Carbajo.



Subscribe to the FRED newsletter


Follow us

Back to Top