Following last week's news that household formation jumped and was revised higher, the logical consequence is that young Americans living in their parents' basement must finally be moving out.

They are not.

In fact, as the chart below from Goldman shows, Millennials are doing anything but moving out, a development that has left Goldman's economists stumped.

Below is a chart showing that the share of young people (18-34) living with parents has held steady over the last half year, and close to the highest since the financial crisis.

This is how Goldman frames its confusion:

“The share of young people living with their parents–which rose sharply during the recession and its aftermath–finally began to decline in 2014. But over the last six months, this decline seems to have stalled.”

But the economy is recovering, jobs are plenty, credit is available to all. How can this be???

Unless… it is all baseless propaganda meant simply to inspire confidence in rigged data.

Unpossible, right? Well, even Goldman is no longer so sure:

We find that the share of young people living with their parents has increased relative to pre-recession rates for all labor force status groups, not just the unemployed and underemployed. Overall, above-average youth underemployment rates alone account for about one-third of the increase in the share of young people living with their parents, and lagged effects of the recession probably account for a bit more.

Goldman tries to explain this counterintuitive… assuming the “intuitive” is that the economy is recovering.

To what extent do current labor market conditions explain the elevated rate of young people living with their parents? To answer this question, we use the CPS micro data to calculate the share of 18-34 year olds living with their parents by labor market status (employed, voluntary part-time, involuntary part-time, unemployed, and not in the labor force). Because the data are noisy and not seasonally adjusted, we use the 12-month average ending June 2015 and then compare with the 2007 average. Our first finding, shown [below] is that the percentage living with parents is higher across all labor force status classifications. Even among the employed, the share of young people living with their parents remains about 2pp higher than in 2007.