Research is not about numbers but rather, telling a story. In family economics it often is about the relationship of household characteristics such as income and education, to investment decisions, retirement adequacy, and general household well being. A frequently used source of data is the Survey of Consumer Finances (SCF) conducted by the Federal Reserve board every three years.The survey is conducted with a large number of randomly selected households. A code book assists researchers by describing the variables (information collected), what they measure, and limitations of those measurements. Because the quality of research and accuracy of the story is dependent on correctly using and interpreting those numbers, we thought it important to see whether researchers in the field are complying with guidelines and suggestions for use of SCF data and pointing out where greater attention to the underlying data is needed. Complex models of analysis and statistics often mask an underlying question of whether the underlying variables and numbers are being used and described accurately.
In exploring this question, we found that many published articles using the Survey of Consumer Finances have not carefully followed proper procedures and guidelines for analysis of the dataset. It is important for researchers and for those reading the research to understand these procedures. There is a large amount of data in the SCF and careful examination of the code book is needed to understand how to use the data.The amount of information for each household participating in the survey can be imagined as being comparable to the size of the household’s federal income tax return: households with low income and assets might have a return of a few pages while households with high assets and business income might have a 100 page tax return.
While there are many questions on attitudes and characteristics of household members that are answered by all respondents, the survey is especially useful for research related to higher wealth households. Out of out of an estimated 505 households in the United States with a net worth over one billion dollars, there might be eight such households in the 2016 SCF.
Researchers and practitioners alike often read articles based on analyses of complex datasets such as the Survey of Consumer Finances. To fully understand and evaluate such articles, it is important that they are aware of details of the datasets and understand what is behind those numbers.
Guest Contributor: Sherman Hanna
Continue the Conversation:
Tell us what you think in the comments below or on Twitter.
Download the research (available to AFCPE members or by request from the authors): Sherman D. Hanna, Kyoung Tae Kim, and Suzanne Lindamood, Behind the Numbers: Understanding the Survey of Consumer Finances, JFCP (29) 2.