Youth Financial Trends 1989-2013

I probed the Survey of Consumer Finances microdata for the period 1989-2013 to determine how the finances of young people have evolved over the period. Here is what I found (spreadsheet).


(Note all of these graphs include all families, including those with zero debt. So they truly reflect the finances all young families as a class.)

As I pointed out previously, median debt levels for families headed by those below the age of 35 are down considerably from 2007

The same is true of mean debt levels as well.

Over this period, the percent of young families carrying debt has remained basically unchanged:

When it comes to the composition of youth debt, this is the picture using means:

Here is the same graph with mortgage debt removed so that you can better see the other trends:

The two big trends in mean debt levels are the sharp rise and then collapse of mortgage debt, which coincides with the housing bubble, and the persistent rise of student debt. Mean student debt carried by these families rose consistently from $1.8k in 1989 to $12.5k in 2014.

I have a hard time knowing exactly what to make of these trends, which is the main reason I find debt analysis difficult. Youth debts are declining primarily because mortgages are declining. But it's not clear that mortgage debt is "bad" debt, except obviously when it's used to secure a home whose value is inflated by an asset bubble. So, in that sense, declining debt may actually be a bad rather than good indicator. Student debt is bad debt to the extent that it reflects increasing higher education unit costs that don't reflect any kind of underlying increase in educational quality. It's also arguably bad debt to the extent that it reflects lower education subsidies. But the rise of student debt is also arguably good to the extent that it reflects the fact that more people are going to college.

If we take the same graph as above and index all the debts to 100 at 1989, in order to get a better sense of percent change over time, it looks like this:

Here is the same graph with student debt removed so you can better see the other trends:

With the exception of student debt, the debt trends all run somewhat similarly. There is a marked run up and high point in 2007, the height of the bubble, and then a subsequent decline. Student debt has, conversely, run up consistently from 1989 straight through 2013.

When we look at what percent of families hold any debt by debt type, we get this:

With the brief exception of credit card debt, less than half of young families has held any given type of debt at any given time. This means the median holdings for each debt type have always been $0.

As far as percentages go, student debt saw the largest rise, with 17.1% of young families holding any student debt in 1989 and 41.4% of young families holding any student debt in 2013. The percent of families with vehicle debt holdings shows no particularly notable trend, except the fall off after the 2007 recession. The percent of young families holding credit card debt has fallen consistently since its high in 1995 (54.6%) to its low in 2013 (36.8%). This too appears to have been sped along by the recession. Finally, consistent with the above graphs, the percent of young families with mortgage debts fell considerably from the 2004 and 2007 highs (37.7% an 37.3% respectively) to the 2013 low (28.6%).

Net Worth

Here is median net worth (assets minus debts) for families below 35:

Here is the same graph for mean:

In both cases, you see the pretty sharp post-2007 collapse, following from the housing bubble. Slight wealth recoveries have begun in the 2010-2013 period.


Here is median family income:

Here is mean:

From 2007 to 2013, median income fell from $41.6k to $35.5k, a decline of 12.3%. Mean income fell during that same period from $58.1k to $48.7k, a decline of 16.2.%


So what is there to learn from this? I am not totally sure, but I can make an effort. If we stick to mean figures for a moment (which summarize all the debt, all the income, and all the net worth), and focus only on the period between 2007 and 2013, we find 1) net worth fell a lot, 2) debts fell a lot, 3) income fell a lot. On first glance, these trends seem quite at odds with one another. Falling debts should boost net worth, all else equal. Falling income should increase debts, all else equal. And so on. The most plausible story I see to put these all together is as follows.

First, income decline is the biggest story. Mean net worth fell $43.6k from 2007 to 2013. Mean income fell $7k fom 2007 to 2010 and another $2.4k from 2010 to 2013. If you average these income declines and multiply them across the 6 years from 2008 to 2013, we get $49.2k. That means that the amount of income lost over these years (roughly derived) approximately equals the amount of net worth lost. The symmetry between the figures is quite neat.

Second, at least some of the decline in mean net worth probably came from falling home values (which hit all families). So one of the reasons the decline in net indebtedness did not translate into higher net worth was because asset value declines were pulling in the other direction.

Third, young families are contending with income declines by reducing their consumption and and holding off on buying homes and to a lesser extent cars. This is part of the reason why an income decline has not translated into higher debt. This also explains the reduction in the percentage of families taking out mortgages.

Fourth, and most speculatively, an increase in student debt levels have caused young families to hold off on buying homes and to a lesser extent cars. This is perhaps one of the reasons why debt levels have not increased even as net worth and income has fallen.

The decline in income (reflected in both mean and median figures) is the easiest major trend that one can identify as unambiguously bad. A 12% (median) or 16% (mean) decline in income is very significant. A $6.1k (median) or $9.4k (mean) annual absolute income decline in income is very large. If these incomes were brought back to their pre-recession levels, it would seem that most of these other trends would clear up quite quickly. For one example of this, consider that the mean student debt increase from 2007 and 2013 was just $3.5k, which amounts to only a few months of the income lost to the recession.