CMD researchers often receive questions about the Consumer Credit Panel
(CCP), the data set underlying the Center’s Quarterly Report on Household Debt and Credit.
Responses to frequently asked questions are organized into the following categories:
Responses to frequently asked questions are organized into the following categories:
Where can I find the microdata? Can I apply for access for my research project?
Access to the Consumer Credit Panel (CCP) microdata is limited to Federal Reserve System researchers, and their coauthors, due to contractual limitations with our data provider. However, we have provided many aggregated looks at the data, which you can find on our data bank.
Your Quarterly Report on Household Debt and Credit only contains data on select states. Where can I find data on all states?
We provide annual data (as of the fourth quarter of each year) on all fifty states plus Puerto Rico, which we typically update during the first quarter of each year. We are unable to provide higher-frequency data on all states due to contractual restrictions with our data provider.
Can you provide data aggregated to even smaller geographic areas?
We do not provide more granular data than at the state level. Sources for more localized geographic detail include the Regional Snapshots on the New York Fed’s public website. For a more national look, our colleagues at the Minneapolis and Philadelphia Federal Reserve Banks have produced the Consumer Credit Explorer.
Can you provide a custom cut of the data?
Unfortunately, we don’t have the capacity to provide custom cuts of the data. Sometimes, we provide analysis and data on a topic that our economists find interesting, but this work is conducted on an ad hoc basis. Still, feel free to reach out to our researchers and express interest—we like to work on topics that engage the public as well.
Do you have information on peer-to-peer (P2P), FinTech, or other types of new lending?
We believe that these types of loans are generally reported on credit reports, and would be captured in the “other” category from our Quarterly Report on Household Debt and Credit. The “other” category contains three subcategories: consumer finance loans (typically unsecured consumer installment debt), retail cards (such as a card that you can only use at a department store), and otherwise unclassified debts. Roughly 50 percent of the “other” category is made up of consumer finance loans, while the remaining half is split roughly evenly between retail cards and otherwise unclassified debts.
How can I get my research paper listed on the website of the Center for Microeconomic Data (CMD)? The paper has been published and uses data that you provided.
We are happy to post a wide array of research employing CMD data on our website and do our best to keep our list up to date. If we’ve missed your work, please let us know, and feel free to do so preemptively.
I’ve used data from a spreadsheet that you provided in a recent report. How can I properly cite the data?
All charts or tables should have the citation “New York Fed Consumer Credit Panel / Equifax.” We provide proper citations in each file, but feel free to ask if you have any questions.
How many people are in your sample?
As of 2018:Q1, there were 13.4 million individuals in our 5 percent primary sample. Our Quarterly Report is calculated using a 0.1 percent primary sample, which comprised 267,000 individuals as of 2018:Q1.The CCP also includes credit reports for individuals residing at the same address as the primary members. Our sampling approach implies a total adult population with a credit report of 267 million people. For more information, please see the 2010 staff report “An Introduction to the Consumer Credit Panel,” by Donghoon Lee and Wilbert van der Klaauw.
Are the data presented in the Quarterly Report on Household Debt and Credit adjusted for inflation?
All data in the Quarterly Report are presented in nominal terms.
Does your information on auto loans include leases?
Yes, the data that we provide on auto loans includes both loans and leases.
Why do the student loan data begin in 2003 when the information on other types of debts begins in 1999?
Student loans did not consistently appear on credit reports until 2003, when the major servicers of federal loans began reporting on both current and delinquent loans. Thus, we do not have reliable information on student loans prior to 2003.
How do the credit card balances that you report handle individuals who charge all of their purchases and then pay them off in full each month?
Our data include the statement balance, which is a snapshot of the most recent statement across all credit card borrowers. It is impossible to distinguish between transactors, or individuals who pay off all of their outstanding balance each month, or revolvers who charge up a large balance and then pay it down slowly.
What happened to the bankruptcy rate in 2005:Q4?
The Bankruptcy Abuse Prevention and Consumer Protection Act went into effect at the end of 2005, causing a spike in personal bankruptcy declarations in anticipation of the change.
What happened to the student loan delinquency rate in 2012:Q3?
In 2012:Q3, there was a reappearance of previously unreported, defaulted student loan debt. These defaulted student loans entered both the numerator and the denominator, causing a noticeable jump in the percentage of student loans that were 90+ days past due. We believe that the delinquency rate before the increase had been understated, and the addition of newly defaulted loans reflects a correction.
What happened to the share of accounts in collection after 2017:Q3?
A change in credit reporting requirements led to a sudden and sharp decline in collection accounts beginning in 2017:Q4. You can read more about the guidance to the credit bureaus here or here for a discussion, or see our “Cleaning Up Collections” post on Liberty Street Economics.
Are “Buy Now, Pay Later” (BNPL) loans included in your report?
Currently, Buy Now, Pay Later loans are likely included in our ‘other’ category in the Quarterly Report if they are reported to credit bureaus, although many are not.
Why don’t your delinquency rates closely align with the delinquency rates that are published by banks or by the Federal Reserve Board?
Our delinquency rates are intended to capture the balance sheet of the consumer, and to capture any difficulties consumers may be experiencing repaying their debts. While bank delinquency rates exclude charged-off balances, they often continue to report those on consumer credit reports as they attempt to collect those defaulted balances from consumers—even after having written them off as a loss. Because those balances remain a responsibility of consumers, we include them in both the numerator and denominator of our 90+-day stock delinquency rate. For a more dynamic look at delinquency rates that excludes the stale, charged-off balances, we suggest the “transitions into delinquency” series from pages 13 and 14 of our Quarterly Report.
Why are the reported debt balances so different than those reported in the Survey of Consumer Finances from the Federal Reserve Board?
For more information on comparing these two data sets, please see “Do We Know What We Owe?” by Meta Brown, Andrew Haughwout, Donghoon Lee, and Wilbert van der Klaauw (2015).
According to the Quarterly Report (page 4), there were approximately 80 million mortgages at the end of 2017, which is much higher than the number reported by other sources. What is the reason for the discrepancy?
The series on page 4 of the Quarterly Report is not adjusted for jointly held accounts (although the figures in the rest of the report are indeed adjusted for double-counting). Thus, this line is an overestimate. We are working to correct this issue.
Why is the balance reported on page 3 of the Quarterly Report so much lower than that reported in the Financial Accounts of the United States / Flow of Funds?
In comparing aggregate measures of household debt to those included in the Board of Governors’ Flow of Funds (FoF) Accounts, there are several important considerations. First, among the different components included in the FoF household debt measure (which also includes debt of nonprofit organizations, private equity, and hedge funds), our measures are directly comparable to two: home mortgage debt and consumer credit. Total mortgage debt and non-mortgage debt were $9.7 trillion and $2.6 trillion, respectively, in 2019:Q3, while the comparable amounts in the FoF for the same quarter were $10.3 trillion and $2.5 trillion. Second, a detailed accounting for the remaining differences between the debt measures from both data sources would require a more detailed breakdown and documentation of the computation of the FoF measures.
Why is the student loan balance reported on page 3 of the Quarterly Report so much lower than that reported in the G.19?
These data sets have two different sources. The data reported in the Quarterly Report may miss some student loan balances that have not yet been reported on credit reports as well as some small share of balances that were defaulted many years before and for which credit reporting has lapsed.
Related Reading
Lee, D., and W. van der Klaauw. 2010. “An Introduction to the New York Fed Consumer Credit Panel,” Federal Reserve Bank of New York Staff Reports, no. 479. November.
Access to the Consumer Credit Panel (CCP) microdata is limited to Federal Reserve System researchers, and their coauthors, due to contractual limitations with our data provider. However, we have provided many aggregated looks at the data, which you can find on our data bank.
Your Quarterly Report on Household Debt and Credit only contains data on select states. Where can I find data on all states?
We provide annual data (as of the fourth quarter of each year) on all fifty states plus Puerto Rico, which we typically update during the first quarter of each year. We are unable to provide higher-frequency data on all states due to contractual restrictions with our data provider.
Can you provide data aggregated to even smaller geographic areas?
We do not provide more granular data than at the state level. Sources for more localized geographic detail include the Regional Snapshots on the New York Fed’s public website. For a more national look, our colleagues at the Minneapolis and Philadelphia Federal Reserve Banks have produced the Consumer Credit Explorer.
Can you provide a custom cut of the data?
Unfortunately, we don’t have the capacity to provide custom cuts of the data. Sometimes, we provide analysis and data on a topic that our economists find interesting, but this work is conducted on an ad hoc basis. Still, feel free to reach out to our researchers and express interest—we like to work on topics that engage the public as well.
Do you have information on peer-to-peer (P2P), FinTech, or other types of new lending?
We believe that these types of loans are generally reported on credit reports, and would be captured in the “other” category from our Quarterly Report on Household Debt and Credit. The “other” category contains three subcategories: consumer finance loans (typically unsecured consumer installment debt), retail cards (such as a card that you can only use at a department store), and otherwise unclassified debts. Roughly 50 percent of the “other” category is made up of consumer finance loans, while the remaining half is split roughly evenly between retail cards and otherwise unclassified debts.
How can I get my research paper listed on the website of the Center for Microeconomic Data (CMD)? The paper has been published and uses data that you provided.
We are happy to post a wide array of research employing CMD data on our website and do our best to keep our list up to date. If we’ve missed your work, please let us know, and feel free to do so preemptively.
I’ve used data from a spreadsheet that you provided in a recent report. How can I properly cite the data?
All charts or tables should have the citation “New York Fed Consumer Credit Panel / Equifax.” We provide proper citations in each file, but feel free to ask if you have any questions.
How many people are in your sample?
As of 2018:Q1, there were 13.4 million individuals in our 5 percent primary sample. Our Quarterly Report is calculated using a 0.1 percent primary sample, which comprised 267,000 individuals as of 2018:Q1.The CCP also includes credit reports for individuals residing at the same address as the primary members. Our sampling approach implies a total adult population with a credit report of 267 million people. For more information, please see the 2010 staff report “An Introduction to the Consumer Credit Panel,” by Donghoon Lee and Wilbert van der Klaauw.
Are the data presented in the Quarterly Report on Household Debt and Credit adjusted for inflation?
All data in the Quarterly Report are presented in nominal terms.
Does your information on auto loans include leases?
Yes, the data that we provide on auto loans includes both loans and leases.
Why do the student loan data begin in 2003 when the information on other types of debts begins in 1999?
Student loans did not consistently appear on credit reports until 2003, when the major servicers of federal loans began reporting on both current and delinquent loans. Thus, we do not have reliable information on student loans prior to 2003.
How do the credit card balances that you report handle individuals who charge all of their purchases and then pay them off in full each month?
Our data include the statement balance, which is a snapshot of the most recent statement across all credit card borrowers. It is impossible to distinguish between transactors, or individuals who pay off all of their outstanding balance each month, or revolvers who charge up a large balance and then pay it down slowly.
What happened to the bankruptcy rate in 2005:Q4?
The Bankruptcy Abuse Prevention and Consumer Protection Act went into effect at the end of 2005, causing a spike in personal bankruptcy declarations in anticipation of the change.
What happened to the student loan delinquency rate in 2012:Q3?
In 2012:Q3, there was a reappearance of previously unreported, defaulted student loan debt. These defaulted student loans entered both the numerator and the denominator, causing a noticeable jump in the percentage of student loans that were 90+ days past due. We believe that the delinquency rate before the increase had been understated, and the addition of newly defaulted loans reflects a correction.
What happened to the share of accounts in collection after 2017:Q3?
A change in credit reporting requirements led to a sudden and sharp decline in collection accounts beginning in 2017:Q4. You can read more about the guidance to the credit bureaus here or here for a discussion, or see our “Cleaning Up Collections” post on Liberty Street Economics.
Are “Buy Now, Pay Later” (BNPL) loans included in your report?
Currently, Buy Now, Pay Later loans are likely included in our ‘other’ category in the Quarterly Report if they are reported to credit bureaus, although many are not.
Why don’t your delinquency rates closely align with the delinquency rates that are published by banks or by the Federal Reserve Board?
Our delinquency rates are intended to capture the balance sheet of the consumer, and to capture any difficulties consumers may be experiencing repaying their debts. While bank delinquency rates exclude charged-off balances, they often continue to report those on consumer credit reports as they attempt to collect those defaulted balances from consumers—even after having written them off as a loss. Because those balances remain a responsibility of consumers, we include them in both the numerator and denominator of our 90+-day stock delinquency rate. For a more dynamic look at delinquency rates that excludes the stale, charged-off balances, we suggest the “transitions into delinquency” series from pages 13 and 14 of our Quarterly Report.
Why are the reported debt balances so different than those reported in the Survey of Consumer Finances from the Federal Reserve Board?
For more information on comparing these two data sets, please see “Do We Know What We Owe?” by Meta Brown, Andrew Haughwout, Donghoon Lee, and Wilbert van der Klaauw (2015).
According to the Quarterly Report (page 4), there were approximately 80 million mortgages at the end of 2017, which is much higher than the number reported by other sources. What is the reason for the discrepancy?
The series on page 4 of the Quarterly Report is not adjusted for jointly held accounts (although the figures in the rest of the report are indeed adjusted for double-counting). Thus, this line is an overestimate. We are working to correct this issue.
Why is the balance reported on page 3 of the Quarterly Report so much lower than that reported in the Financial Accounts of the United States / Flow of Funds?
In comparing aggregate measures of household debt to those included in the Board of Governors’ Flow of Funds (FoF) Accounts, there are several important considerations. First, among the different components included in the FoF household debt measure (which also includes debt of nonprofit organizations, private equity, and hedge funds), our measures are directly comparable to two: home mortgage debt and consumer credit. Total mortgage debt and non-mortgage debt were $9.7 trillion and $2.6 trillion, respectively, in 2019:Q3, while the comparable amounts in the FoF for the same quarter were $10.3 trillion and $2.5 trillion. Second, a detailed accounting for the remaining differences between the debt measures from both data sources would require a more detailed breakdown and documentation of the computation of the FoF measures.
Why is the student loan balance reported on page 3 of the Quarterly Report so much lower than that reported in the G.19?
These data sets have two different sources. The data reported in the Quarterly Report may miss some student loan balances that have not yet been reported on credit reports as well as some small share of balances that were defaulted many years before and for which credit reporting has lapsed.
Related Reading
Lee, D., and W. van der Klaauw. 2010. “An Introduction to the New York Fed Consumer Credit Panel,” Federal Reserve Bank of New York Staff Reports, no. 479. November.
