With all the discussion about today’s savings rate, it’s important to point out an important fact. Savings is a residual: That means the government measures disposable income and consumption spending. Then they subtract the second from the first, and what’s left is savings.
The implication is that any revisions in either income or consumption immediately and directly affect savings in a big way. For example, suppose that the income of a family is measured as $100 and its consumption is measured as $95. That leaves savings of $5, or a 5% rate.
But if measured income is revised up to $105—a 5% increase—then savings will be revised to $10, and the savings rate will nearly double to 9.5%.
What does this have to do with the real world? Let’s take a look. First, here is the familiar chart of the personal savings rate, as currently reported by the BEA:
In this chart, the savings rate takes an inexorable plunge down, from roughly 11% in 1981 and 1982 to almost nothing in recent years. Oh, those early Americans were such good savers…
But wait! Here’s another chart, of the personal savings rate as originally reported. That is, for each year I used the savings rate as reported in the Survey of Current Business, the February or March issue of the next year (i.e. I got the 1981 savings rate from the February 1982 SCB).
Hmph…all of a sudden that previous generation doesn’t look quite so prudent. The savings rate in 1981 is reported today as 10.9%, but at the time it was originally as 5.3%. Or take 1986—today it looks like the savings rate was a strong 8.2%, but at the time it was originally reported with a savings rate of 3.8%, less than half. In fact, through most of the early 1980s, journalists and economists were bemoaning the low savings rate, based on the numbers they were seeing.
Now, I’m sure Steve Landefeld and his friends at the BEA had very good reasons for revising the statistics (hi Steve!). But it’s worth noting that there are some statistics—and the savings rate is one of them—which are highly prone to large revisions.
On the next page, I put both lines on the same chart.
This chart shows the current savings rate data versus as reported.
Note: For these charts I used the February or March data immediately following the year. But if I wait an extra year, the same pattern remains. That is, the savings rate for 1982 was originally reported as 6.6% in early 1983. That estimate for 1982 was actually lowered to 5.8% by early 1984.