The Dataset: Distributional Financial Accounts (DFA) [1] – Share of Aggregate Household Wealth by Education
Information Included in the Dataset
Starting with the general organization of our dataset, the data is grouped by rows of years and quarters. For example, the actual data stretches from Quarter 3 of 1989 to Quarter 3 of 2025. The columns with financial data all correspond to a specific year and quarter.
Moving on to the specific financial data present in each row, the columns organize the data to most notably include:
- Net worth
- Assets
- Loans
- Financial Assets
- Non-financial assets
- Real estate
- Home mortgages
- Deposits
- Corporate equities and mutual fund shares
- Household count
Related fields to these financial categories are included in the dataset columns as well; the critique details the most notable ones.
Each of these fields, other than household count, is expressed as a percentage of a category, being:
- College
- SomeCollege
- HS
- NoHS
in education levels.
Example
Information, Events, or Phenomena the Dataset Illuminates
Our dataset illustrates how different financial outcomes in the United States are distributed across different levels of educational attainment from 1990 to 2024. By examining variables such as total assets, real estate, and other forms of equity, the data can reveal patterns of economic inequality in the United States and how education levels play a pivotal role in achieving financial security.
The dataset can also illuminate levels of social mobility and how those with a higher level of education have more access to wealth-building opportunities and overall more economic stability compared to those with a lower level of education.
In addition, the dataset can demonstrate long-term trends, demonstrating how an increase of education attainment in the U.S. from 1990 to 2024 is also correlated with an increase in financial success across the population.
What the Dataset Cannot Reveal
The most significant limitation of our dataset is that it oversimplifies education into a binary metric, completely overlooking the differences in quality. For example, DFA cannot reveal the disparity between a local public University and an Ivy League school.
Furthermore, the data treats wealth as a passive accumulation, completely neglecting the physical and mental effort, grueling work that one had to endure on their way to success. The dataset is also not able to distinguish whether someone’s wealth came from their own job or that it was handed to them by their parents.
How the Data was Generated
The DFA data were generated and collected by the Federal Reserve, the Financial Accounts of the U.S., and the Survey of Consumer Finances (SCF). The Financial Accounts provide quarterly totals for household assets and liabilities across the entire United States economy. The SCF is a detailed household survey conducted every three years that shows how wealth is distributed by characteristics like education, income, and age.
To create the DFA, the Fed takes into account differences between the two datasets, then uses SCF data to estimate wealth shares for different groups. The estimated shares are then forecasted to create seasonal estimates. The shares are then applied to the aggregate totals from the Financial Accounts, producing consistent quarterly distributional wealth data, which is then applied to the measurement of different social structures, like education.
What Information is Left Out & Dataset’s Ontology
The dataset presents all financial variables as percentage shares rather than absolute dollar amounts, which obscures the actual magnitude of wealth held by each education group. The data also lacks demographic breakdowns beyond education level, as variables such as age, income, race, gender, and geographic region are absent. This prevents researchers from investigating how education intersects with other socioeconomic factors in shaping wealth distribution.
Additionally, the four education categories, from high school to college, do not distinguish between bachelor’s and advanced degrees, nor the fields of study, limiting analysis of how different educational paths affect wealth outcomes. Also, there could be institutional factors, such as financial literacy, access to banking services, or employee benefits, that are not represented, which are part of their either active or passive income.
The spreadsheet’s ontology is constructed around the spreadsheet columns we directly see. Overall, financial measures are highlighted, but other factors, such as stable housing and access to public services, are left out. Additionally, the spreadsheet’s ontology fails to detail the uncertainty of the data presented on it.
The sources underlying the DFA dataset are the Z.1 Financial Accounts and the SCF, both produced by the Federal Reserve. The Z.1 accounts gather data from various administrative and institutional sources, including regulatory filings and reports, to measure aggregate levels and flows of financial assets and liabilities across sectors of the U.S. economy. These accounts are updated each quarter, providing comprehensive quarterly measures for the household sector, but do not include demographic detail.
In contrast, the SCF is a triennial household-level survey that directly collects information on assets, debts, income, and demographic characteristics such as education, offering detailed distributional insight but at a lower frequency. The survey is collected by NORC at the University of Chicago.
The DFA dataset was constructed and funded by the Federal Reserve Board, which is the central banking system of the United States. Because the Federal Reserve is a public institution, the dataset is publicly funded rather than sponsored by private companies.
The DFA project builds on existing Federal Reserve data products, such as the financial accounts and the Survey of Consumer Finances, which are also produced and funded by the Federal Reserve. The research behind the DFA is documented in Federal Reserve publications such as FEDS Notes and working papers. Overall, the dataset reflects a government-funded effort to improve knowledge and understanding of wealth inequality in the United States.