Didn't happen:
The above table (click on it for a larger, more readable version) uses the latest BEA data, which was updated earlier in 2010. As you can see, personal savings is now well above zero for 2005. FYI -- when the BEA updates its data, it overwrites the old data.
I discovered this feat of prestidigitation while reviewing a 2007 Fed paper shortly after I generated the above table, which I redrew to put it on the same scale as the below chart:.
(click for larger image)
The black line shows personal savings as reported by the BEA at the time of publication of this paper back in 2007. You can see that it actually dives below zero in 2005.
How did the BEA accomplish this? Simple, they inflated the data with "imputations," something that John Williams and Mike Shedlock have discussed here and here, respectively. Essentially, BEA economists use imputations to impute non-monetary benefits (e.g., meals paid for by your employer) as money in your pocket.
So, what can imputations do for you? Check out this drive:
(click on image for larger chart)
You can see from comparing the RAW savings rate data to the data used by the Fed paper that even that data had been inflated with imputations, only less so. The fact is, without imputations, our savings in the aggregate went negative in 1998!
In the future I will have a lot more to say about imputations and other stupid accounting tricks that the BEA, BLS and the Federal Reserve rely upon overstate economic growth and understate unemployment, and I will tie this all back to the entrenchment of neoliberalism in the federal government beginning with the closing of the gold window back in 1971. Should be fun.
