BARRON'S COVER

The Great American Savings Myth

By GENE EPSTEIN

IF YOU BELIEVE THE OFFICIAL NUMBERS, personal saving in the U.S. has collapsed: It fell in the late 1990s, tumbled all the way to a 70-year low by 2001, then cascaded off that cliff into minus territory in 2005 and has languished there ever since. It seems like an impending catastrophe. If Americans live beyond their means they can't accumulate the wealth to provide for their retirement, much less create a nest egg big enough to see them through rainy days.

But a look at that nest egg calls into question the whole personal savings story. Household net worth -- assets minus debt -- has never been higher, having grown rapidly even as the personal savings rate nose-dived (see chart below). How can we be getting richer without saving? It's more likely that the existing definition of personal savings will no longer do. In fact, a broader measure, far from running negative, reached a 50-year peak in 2004 and was still near it by 2006 (lower chart).

Americans may still have to save more to cover the current and future debts incurred by their profligate Uncle Sam. And aging baby boomers might have to save more to meet their retirement objectives. But in historical terms, at least, saving in the U.S. is alive and flourishing.

How can our finding differ so much from the official story? The key reason is that the mix of savings has shifted toward flows the official data don't capture. As Barron's just last week pointed out ("Still Too Stingy,¹" May 21), dividend payouts are at a record low. The cash withheld from shareholders -- and then some -- is invested by companies in assets that are tangible such as plant and equipment and assets that are intangible such as brands, patents and licenses.

Since both forms of investment are aimed at boosting the value of a company's stock, both enhance a shareholder household's net worth. Equity in all business accounted for 30% of this net worth by year-end 2006. Yet for the most part, neither type of investment is reflected in the official data on personal savings.

---

---

Spending on intangible assets is particularly huge and rapidly growing. Baruch Lev, a New York University finance professor who has written extensively on this subject, calls them "today's major value driver of business." Yet, with the exception of computer software spending, these investments are omitted from even the broader government figures on savings in the national income accounts.

PEOPLE OFTEN ASSOCIATE "SAVINGS" WITH THE quaint picture of a hard-working employee dutifully depositing part of a paycheck in a passbook bank account. Such venues are hardly extinct. Of the $55.6 trillion in household net worth by year-end 2006, $6.7 trillion was in checking accounts, time deposits and money-market funds. But in a modern economy, most saving is money put at risk, often with the hope of large returns and the chance of substantial losses.

So even if the dramatic rise in net worth is based on asset bubbles waiting to burst -- though the historical evidence belies this dour view -- it would not alter the fact that saving occurred. It would mean only that the saving was misallocated.

What's most important to understand is that saving increasingly happens through a type of automatic withholding -- without the saver's conscious participation. Baby boomers may never have acquired the virtue of thrift that came naturally to their parents' generation. More likely, their relative success at saving results from the greater prevalence of institutional aids serving them in their capacity both as shareholders and -- as we'll see -- homeowners.

Let's look at some of the numbers to see how record household riches can occur despite an anemic -- at best -- official savings rate. Household net worth -- assets minus liabilities -- stood at a record $55.6 trillion by the end of 2006, according to "Flow of Funds" data from the Federal Reserve Board. Owing to methodological limitations, the net-worth figures also include assets and debt held by nonprofit organizations that are hard to separate from the totals.

But Fed statisticians still find the household totals accurate enough to run them against the standard measure of disposable personal income, which is after-tax income. Household net worth was 584.1% of DPI by year-end 2006, the second highest multiple on record. To arrive at real-net-worth per household (seen in top chart), annual data from 1956 through 2006 were first inflation-adjusted in terms of 2006 dollars, then calculated per household by dividing each figure by the growing number of households. For example, since prices rose by 22% from 1996 to 2006, net worth had to rise by 22% just to remain the same in real terms. Since the number of households rose by 15% over that period, real net worth had to rise by 15% to stay the same per household.

The result of these two adjustments puts real net worth per household at a record high of $486,000 by year-end 2006, or 31.7% higher than at 1996's conclusion, one of the fastest 10-year increases ever.

But can't this figure plummet if asset values fall as they have in the past? What about the precarious state of many housing markets? Yes, they can fall, but the effect can be offset by other household assets.

Assume the value of homeowners' equity falls by as much as a third -- an unprecedented setback given that the biggest-decline ever in homeowners' equity ran a little over 12% back in 1974. That kind of drop would punish disproportionately those households whose major assets were their homes.

But since equity in a home made up just one-fifth of total household net worth as of year-end 2006, the overall impact would be muted: The total household decline would be less than 7%. Even after such a big hit, household net worth would still be 22.9% higher than in 1996.

If you look at the 50-year trend in real net worth per household against the conventional measure of personal saving (top chart), the results seem strange. Saving often looks strong just when gains in net worth are weak -- and weak just when gains in net worth are strong.

WHATEVER THE VIRTUES OF HOARDING FOR ITS OWN sake, saving is more properly viewed as a necessary sacrifice for the accumulation of assets; try adding a "y" to "miser." So if those two conflicting trend lines were a given, it would be only proper to celebrate the amazing rise of the Free-Lunch Economy, as we rejoice in the spectacle of Americans building wealth by saving nothing. Or we could question whether something is missing in the official data on personal saving.

A lot is missing in the official data. Calculated by the Bureau of Economic Analysis, or BEA, the Commerce Department agency in charge of the national income accounts, "personal saving" is defined in a way that sounds plausible. It is essentially whatever is left over after spending on consumption is subtracted from disposable personal income.

But to begin with, for reasons that may be defensible but are certainly controversial, DPI specifically excludes all income realized from capital gains that might otherwise go into saving. For example, say an investor takes a capital gain of $20,000 on the sale of IBM shares and pays $3,000 in taxes on it. The $20,000 does not get added to pre-tax personal income, but the $3,000 tax does get subtracted to arrive at DPI. The $17,000 difference might have been saved but it is not part of the calculation.

BEA staff economist Marshall Reinsdorf has shown that recent negative savings figures become positive merely by restoring capital gains taxes to income. But Reinsdorf still finds ("Alternative Measures of Saving," Survey of Current Business, February 2007), that these and other limited fixes do not change the long-term downtrend in the savings rate by very much.

To repair the disconnect between savings and wealth, it is first necessary to recognize that institutions don't save, only people do, even if they don't do it consciously the way the previous generation did.

Far from running negative in 2006, personal savings, under our broader definition, came to $2.66 trillion -- a record high in nominal dollars. The lower chart applies the same adjustments to the savings figures previously applied to household net worth, with all figures since 1956 put in 2006 dollars and then divided by the number of households. On that basis, real saving per household has been on a plateau over the past several years, peaking at $23,800 in 2004 and still running at $23,250 in 2006.

The savings totals for each year are drawn from two main sources. First, the BEA regularly estimates "gross private saving," which captures all contributions to saving from private sources, excluding the much-touted saving that comes in from abroad. For example, out of the $2.66 trillion total in 2006, $1.72 trillion came from gross private investment. (See "Table 5.1. Saving and Investment," at www.bea.gov.)

This figure reflects all private contributions to investment in plant and equipment by business. These investments, in turn, support gains in equity value in both publicly traded and closely held businesses -- which, if you'll recall, accounted for about 30% of total household net worth by year-end 2006.

Gross private saving further includes contributions to the building of new residences and the renovation of existing homes, which can help power gains in the value of homeowners' equity. These forms of residential investment far exceed the sums imputed to homeowners in the BEA's conventional measure of personal saving. As noted, by year-end 2006, homeowners' equity accounted for about 20% of total household net worth.

Spending on computer software is the one major investment area covered by gross private saving that relates to intangible assets. The BEA began incorporating this form of investment in its national income accounts by the 1990s.

But until the agency goes further, estimates are needed for gross spending on intangibles. Out of the $2.66 trillion of gross saving in 2006, this spending ran to $0.94 trillion, based on estimates prepared for Barron's by Philadelphia Federal Reserve economist Leonard Nakamura. (See his essay, "A Trillion Dollars a Year in Intangible Investment and the New Economy" in Intangible Assets, edited by John Hand and Baruch Lev, 2003.)

Economists have long recognized that intangibles like patents in prescription drugs, copyrights in books and articles, licenses in software, brand names, trademarks, trade secrets and unique organizational structures can often be far more valuable than tangible corporate assets. But their creation isn't cost free. Saving is required to create them just as surely as it is needed to create physical assets.

Nakamura not only includes the surge in the number of dollars allocated to conventional research and development, he also includes the resources devoted to the decision-making process when new products are being introduced, plus the money spent on advertising.

Since intangibles are the fastest-growing form of investment, its relative neglect is a key reason why personal saving is so vastly understated. BEA economist Reinsdorf likens the attempt to reconcile the gains of household net worth with real personal saving to the research on the real sources of productivity gains in the late-1990s.

"Economists began to realize," he observes, "that faster productivity growth could not just be due to investment in plant and equipment. It was also driven by the additional effort of investing in intangibles. In this case also, much more than traditionally defined personal saving is required to explain the rise of net worth."

The lesson in both cases? There is no free lunch.