Some people are worried about high levels of U.S. household debt. When looking at aggregate numbers, there are two ways to consider this question. The first is how much it costs to service this debt as a fraction of disposable (after tax) income. This is shown with the blue line. The second is how much debt there is with respect to the same disposable income measure. This is shown with the red line. Whether these numbers are high is difficult to say; household-level data are more appropriate for that question. But in the aggregate, both measures have clearly decreased during the past crisis. Note the scale, though: While service payments decreased by almost one-third, the debt ratio decreased by only one-fifth. And whenever interest rates go back up, service payments will increase.

How this graph was created: Creating the blue line is easy: Search for “household debt” and select the series for debt service as percent of disposable personal income. The red line is more complex because it has to be constructed: We need the two components of household debt (consumer credit and mortgages) as well as nominal disposable income—nominal, not the real or per capita versions, because the debt measures are in nominal terms. So, from within the graph, search for “household consumer debt” and add this series (a) to the graph. We must combine more data here, so add “household mortgage debt” (b) and “disposable income” (c), being sure to select “modify series 2.” Then create your own data transformation by applying the formula (a+b)/c. Finally, switch the y-axis position to the right.

Suggested by Christian Zimmermann