Bloomberg Anywhere Remote Login Bloomberg Terminal Demo Request


Connecting decision makers to a dynamic network of information, people and ideas, Bloomberg quickly and accurately delivers business and financial information, news and insight around the world.


Financial Products

Enterprise Products


Customer Support

  • Americas

    +1 212 318 2000

  • Europe, Middle East, & Africa

    +44 20 7330 7500

  • Asia Pacific

    +65 6212 1000


Industry Products

Media Services

Follow Us

A Savings Puzzle

Posted by: Michael Mandel on June 10

I am a connoisseur of statistical discrepancies—and I’ve found a big one in the savings data that I still don’t understand. This is going to be extremely wonky—you are forewarned.

We all know that the personal savings rate has stayed remarkably low, despite the slump in housing prices. In the first quarter, the BEA figures show that personal savings was a tiny 0.6% of disposable personal income, barely higher than it was in 2005-2007.

The Federal Reserve also calculates a gross household savings number, which takes personal savings and adds in investment in consumer durables and consumption of fixed capital (depreciation). This measure, too, has not risen very much. (Note: The household numbers from Table F.100 include nonprofits as well).

But here’s the thing. The Fed, in its flow of funds release, calculates a number for gross household investment. Gross household investment is the combination of household capital spending and net financial investment. That is, the amount you save can either be used on physical investment goods (a house, a car), put into a financial asset (a savings account, money market funds, stocks, or so forth), or used to pay down debts (mortgages, credit card debts).

In theory, gross household savings should equal gross household investment—but it doesn’t. In fact, gross household investment has been soaring. Take a look at this chart.


The blue line is gross household savings, the red line is gross household investment. In theory, these should be the same lines. But in fact, the statistical discrepancy—the excess of savings over investment—has swung from roughly $300 billion in 2005 to -$390 billion in the first quarter of 2008 (at annual rates).

To put it another way, even though the official savings rate is low, Americans have supposedly been funnelling massive amounts of money into financial assets. Net financial investment has gone from -741 billion in 2005 to $219 billion in the first quarter of 2008 (at annual rates). See this chart:


In other words, Americans who were borrowing heavily in 2005, 2006, and 2007 are now lending again (that’s what the shift from negative to positive means).

What are the possible explanations for this shift?

1) Just a fluke of the data—many statistical discrepancies end up being revised away as new source data comes in. So it may turn with this one as well.

2) Personal savings could be really higher than the numbers show—either because consumption is lower or because personal income is higher. There is in fact a long history of upward revisions in the personal savings numbers. If consumption is lower, then the economy might be weaker than it seems. Or if personal income is higher, it could be stronger.

3) The financial data could be mistakenly attributing investment by businesses and foreigners to American households. In fact, since the foreign financial flow data is squishy, there may be more money coming into the country, net, than the statisticians realize.

So no answer here…just questions.

TrackBack URL for this entry:

Reader Comments


June 10, 2008 02:41 PM

In an open economy (like a household) investment doesn't have to equal savings (at least in the short run).

Mike Mandel

June 10, 2008 03:19 PM

I think that these household savings and investment numbers are supposed to be equal. The "rest of the world" is a separate sector


June 10, 2008 08:06 PM

Without looking at the raw data, I'm assuming that the debt repayment is simply a residual calculation. That is, the Fed is not calculating the sum of the payment made, but rather the differences in the debt. So foreclosures and other debt write-offs will have the same effects as pay-downs. That's why we're screwed.


June 10, 2008 11:22 PM

Very interesting post Michael...
The answers could be in a detailed analysis of how all these statistics are calculated and the amount of "data massaging" applied.
I confess my ignorance in the matter, I do not have a detailed knowledge of all the "tweaking" applied in this particular it could be a fluke.
It could be N. 3 first look it seems to me the best candidate.
Or, as you suggested, this data campletely fails to capture some important "piece of reality" in the computation.
I know, I know I stated the obvious and I didn't give you any specific answers...sorry ;-)


June 11, 2008 05:10 AM

You may have omitted the holding gains on assets from table R100 in the FOF accounts. You need both F100 and R100 to reconcile NIPA and FOF saving.


June 11, 2008 10:30 PM

I have read a few analytical pieces over the years that outlined how the household savings data were deeply flawed and mismatched. It is a residual calculation where income less taxes, less spending equals savings.

However, not all income is captured (401k contributions and capital gains are not included as income), while the taxes on all income (including capital gains) is subtracted. And some investment-like expenditures are counted as spending. Accordingly, I would not expect the household savings data to tie to anything, nor to have any relevance to household wealth accumulation.

As capital gains increase and people make greater use of 401k accounts the calculation of household savings rate should decline. I believe we saw this during the 1990s.

So it is possible to have a negative household savings rate while our wealth grows.

June 13, 2008 01:26 AM

I believe you'll find what you're looking for in the way that the data sources deal with the auto loans, home loans, and home equity loans, the acquisition of debt that is large relative to income, savings, and investment).


July 14, 2008 04:19 PM

This discussion highlights a major issue with our economic system today: even intelligent analysts can't make sense of the data that the government is providing on our economy. How can we drive the car when we have no idea what speed it's going or if the steering wheel actually will point the car in the right direction? Economic statistics shouldn't be "calculated", they should be simple sums of simple numbers. Anything beyond that introduces measurement error, which distorts our ability to perceive reality. In fact, I think it's high time to revisit our private and public sector reliance on the art of statistics. It's pretty clear that S&P and Moody's screwed up their ratings because of statistics. Now it looks like our economy is being screwed up because of the same.
If you can't sample it correctly, then do the hard work and count them.

Thank you for your interest. This blog is no longer active.



Michael Mandel, BW's award-winning chief economist, provides his unique perspective on the hot economic issues of the day. From globalization to the future of work to the ups and downs of the financial markets, Mandel-named 2006 economic journalist of the year by the World Leadership Forum-offers cutting edge analysis and commentary.

BW Mall - Sponsored Links

Buy a link now!