The shifts in household debt and the need for more data

The Federal Reserve Bank of New York released data yesterday on household debt and credit for the second quarter of 2014. The data show that total household debt dropped by 0.2 percent, or $18 billion, from the first three months of 2014. So the amount of debt was essentially unchanged at $11.63 trillion. To put that figure in perspective, the ratio of household debt to gross domestic product in the second quarter was about 67 percent, which is lower than 85 percent, the ratio in the third quarter of 2008 just before the financial crisis.

This tells us a lot about the state of our economy today, but it also highlights the kind of data we need to more fully understand whether and how debt is affecting economic growth and inequality in the short-term. But first, let’s look at what the data released yesterday do tell us.

Overall household debt was essentially unchanged, but shifts in the composition of debt are obvious. Mortgage debt, about 70 percent of all household debt, declined during the second quarter continuing a post-crisis trend. According to the New York Fed, the amount of mortgage origination, essentially new mortgages, declined to the lowest level since 2000. And not only are new mortgages at a low level, but so are problems with mortgages. According to the data, the share of mortgages going into delinquencies was also at a level not seen since 2000. Households appear to be steadily recovering from the household boom and bust of the last decade and are less willing to enter into the mortgage market.

But other kinds of household debt are increasing. In particular, student loans increased moderately during the quarter, but have jumped sharply over the past year. Loan balances increased by $7 billion to $1.12 trillion, or about 10 percent of all household debt. Over the past year, student loan debt has increased by about 12 percent—and households are having a hard time servicing. About 10 percent of student loan debt is delinquent or in default for 90 days or more. The overall economic consequences of student debt aren’t yet fully understood, but these data aren’t a good sign that households are handling the increase well.

The dataset also contains data on auto loans, a source of consumption growth for the overall economy. Economists at the New York Fed dug into the data yesterday looking particularly at subprime auto loans, or loans going to borrowers with low credit scores. They found that while subprime loans are below pre-recessions levels, they are increasing rapidly. And they found that the source of these loans isn’t traditional banks, but auto finance companies.

These data are interesting and useful for our understanding of the economy, but they only present aggregate figures on debt and its composition. These data do not allow economists and other analysts to look at quarter-on-quarter growth in different kinds of debt by income or wealth levels. The Survey of Consumer Finances presents data like that, but it’s only available every three years.

Higher frequency data would give us a better understanding of not only the health of the macroeconomy, but the changing distribution and composition of household debt. This is something policymakers in the Obama administration and Congress should take a hard look at enabling.