The United States is being held hostage by a dubious statistic and a serious misapprehension. The statistic shows that household saving in the U.S. economy dropped precipitously during the 1980s. The serious misapprehension is that this drop has impaired economic growth and that the economy cannot revive until the savings rate increases. In the standard view, without savings there can be no investment, and without investment, no growth -- whence comes the deceptively simple but misleading idea that the path to recovery lies in a revival of household savings.
The drop in household saving over the past decade reported by the Commerce Department is certainly dramatic. From 7.5 percent of disposable personal income in 1981, saving fell to only 2.9 percent in 1987, the lowest ever recorded. Since then the rate has moved up to just over 4 percent.
Household saving by no means comprises total national saving. It is dwarfed by corporate saving (profits plus depreciation), which is some five times as large. Corporate saving shows no downward trend; it amounted to 12.8 percent of gross national product (GNP) in 1981 and 12.4 percent in 1988.
The savings rate is also often said, mistakenly, to be affected by the "drain" of government deficits on private savings. However, the failure of government accounting to differentiate government investment from government consumption makes it impossible to know whether public deficits are using private savings to finance consumption or investment. Other advanced nations manage to combine public-sector deficits of 3 or 4 percent of GNP with ample private savings rates. Thus, with corporate saving steady, the real debate over saving rates in the U.S goes back to whether household saving is down. Household saving has also received special attention because its reported decline is taken as evidence that Americans are failing in their personal responsibility for prosperity. Indeed, the drop from 7.5 percent to 2.9 percent of disposable personal income does suggest an alarming decline in our collective provision for the future.
But is this decline to be believed? The answer depends entirely on how we calculate it. The Department of Commerce, which prepares the National Income and Product Accounts that are the official books of the economy, does not directly measure how much the nation saves. Instead, it calculates saving in two steps. First, it measures "disposable personal income," which is total household income minus tax payments. Then it subtracts "personal outlays," a measure of household spending on consumption that includes all personal goods and services, except houses, which are counted as investment. The difference between disposable personal income and personal outlays is, in this approach, what households save. For the second half of 1991, for example, disposable personal income ran at an annual rate of $4,068 trillion and personal outlays at $3,898 trillion. Personal saving was the difference, or $170 billion. This figure, divided by disposable income, yields our "personal" saving rate of 4.2 percent.
While there is nothing inherently wrong with measuring saving this way, it has one serious drawback. Because saving is the rather small difference between two large numbers, small proportional changes in the larger numbers greatly affect the result. For example, if total income for 1991 was underestimated by 4 percent, while estimates of consumption spending were accurate, savings would jump from $170 billion to $332 billion, and the saving rate from 4.2 percent to 7.8 percent, right up there in the big leagues. In our view, this is exactly what happened in the 1980s; indeed, the Commerce Department underestimated income by much more than 4 percent. When we take that error into account, the much touted decline in the saving rate disappears.
The Capital Omission
We need to make three adjustments to the Commerce Department figures to bring them in closer touch with economic realities. The first is a proper accounting of capital gains.
Disposable personal income, as the Commerce Department measures it, currently does not include the capital gains that come to individuals when they sell stocks or other assets. There has been a long debate as to whether or not these gains are properly considered income, since they do not reflect any earnings from additional output but only changes in prices. We do not need to challenge the Commerce Department's definitions, which seek to portray the real changes in the economy. But in calculating the financial surplus available for investment, it is clearly a mistake to omit capital gains. Not only are capital gains a source of much business financing; during the 1980s, they reached unprecedented levels.
A look back to 1970 makes the point. Capital gains that year amounted to $21.3 billion, or 3 percent of disposable income. In the 1980s, they grew explosively, from $70.8 billion in 1980 to a peak of $295.8 billion, or 9.8 percent of disposable income, in 1986. The 1986 level was exceptionally high because the tax reforms adopted that year led investors to take capital gains while they still enjoyed lower tax rates. Yet even the 1987 and 1988 levels of capital gains were high by historical standards.
No plausible measure of the national saving rate should ignore this immense addition to the financial investment power of households, or more accurately, of those households at the apex of the income pyramid. Moreover, these figures are conservative because they omit another $60 billion to $80 billion a year of capital gains income that did not have to be reported to the Internal Revenue Service, including most gains on sales of owner-occupied homes.
For some economists, the idea of including realized capital gains in household income is heretical. However, dollars earned by households from asset sales do not disappear just because the national income accounts exclude them from income. Moreover, the accounts do treat as income all capital gains earned on assets held for less than six months, while omitting gains on assets held longer. This exclusion becomes a problem in the measurement of savings when long-term capital gains increase as a proportion of personal income -- and when the statistics fail to reflect the change.
Missing Pension Funds
A second problem arises in the treatment of net household contributions to pension funds -- that is, any excess of contributions (or earnings on past contributions) over what pension funds pay out. Pension plans are a critical part of the saving story. For all but the richest 20 percent of households, most financial saving occurs through pension plans. Only the highest-income households save significant amounts by accumulating stocks, bonds, and bank accounts. And since pension assets grew rapidly during the 1980s, there is no basis for the view that spendthrift middle-income Americans caused a savings shortfall.
In the National Income and Product Accounts, net contributions to private pension funds are counted as part of household income and therefore are figured into household saving. But when the net contributions go to state and local government pension funds or to the Social Security TRust Fund, they are not counted that way. These accounting decisions, like the classification of capital gains, do have a rationale. But, again, accounting procedures need to reflect changes in the real world. In 1975 household net contributions to state and local government pension plans amounted to only $7 billion; they reached $69 billion in 1988 and are still growing. A similar shift took place in Social Security contributions. Prior to 1984, net household contributions were negative; families were taking more out of the Social Security funds than they were putting into them. But in 1984, as a result of changes adopted by Congress, the flow was reversed. In 1988 a net flow of $53 billion went from household incomes into the Treasury bonds in which Social Security TRust Funds are invested.
Should we count these very large net contributions as saving? The Department of Commerce ignores contributions to public pension funds on the grounds that they are spent to finance "deficits," not "investments," as are private pension funds. One reason why the Commerce Department makes this determination is that it fails to divide the government deficit into consumption and capital accounts. Since it does not officially recognize a category of public investment against which it could credit government pension contributions, the Commerce Department finds no alternative but to consider them as financing mere consumption, and therefore not deserving the accolade of saving.
Of course, state and local pension funds invest in the capital markets, and Social Security purchases of Treasury bonds go into the same pool as capital from private buyers that also purchase those bonds. Yet, coming from government, these funds are counted as consumption; coming from private sources, they count as investment. Once more the data should not ignore a remarkable change in the pool of funds for financing investment. Taken together, public net pension flows amounted to less than 1 percent of disposable income in 1980. By 1988 they had more than tripled; in that year household contributions increased government pension reserves by $122 billions. Since we treat contributions to private pension funds as saving, it seems illogical to exclude contributions to public pension funds, especially because they increased so markedly during the years when the saving rate underwent its mysterious decline.
The Effect of Housing Inflation
The final source of the great saving decline arises from the official treatment of owner-occupied housing. Here the Commerce Department's division of personal outlays into consumption and investment enters into the saving controversy.
The Commerce Department calculates the amount that households allow for wear-and-tear on their housing, the personal equivalent of the set-aside that businesses take as the depreciation on their plant. This expense is then subtracted from the income that these owner-occupants enjoy as "free rent." Of course, no one we know, including our economist friends, figures annual income that way. But, as before, the important point is the effect of accounting procedures on the official measure of saving.
During the 1980s, real estate was rapidly inflating. As a result, the depreciation calculated by the Commerce Department also increased sharply. As this depreciation "expense" grew, the presumed income of these households diminished, whether or not they were aware of it.
The impact of this accounting illusion was substantial. If the imaginary depreciation charge had not been levied against household income, savings in 1988 would have been measured at $222 billion, rather than $145 billion, raising the saving rate from 4.2 to 6.2 percent.
Adding Up the Errors
It remains only to sum up. The bars in Figure 1 show the estimates for personal saving from the Commerce Department's National Income and Product Accounts (identified as "NIPA"), along with the dollar adjustments from adding to householders' incomes the net proceeds of their capital gains ("Cap Gains"), contributions to government pension funds ("Pensions"), and depreciation expenses on their housing ("Housing"). Figure 2 shows the unadjusted and adjusted NIPA saving rates.
One major conclusion leaps out. It is the complete disappearance of any fall in our saving performance, once allowance is made for changes in the real economy whose effects did not show up in the statistics. The most important of these changes, by far, was the Wall Street frenzy that generated the surge of capital gains, which financed perhaps half to two-thirds of all household surplus during the decade. Higher capital gains, together with the statistical side-effects of real estate inflation and growing public pension plans, created a serious gap between the measurement of saving in the national income accounts and the real-world capacity to finance investment. The official index showed a serious decline in household saving. A properly adjusted measure shows none. Corroboration from the Fed
Another reason to prefer our adjusted measure of saving is the testimony of an entirely separate source -- data compiled by the Federal Reserve Board. Unlike the Department of Commerce, the Federal Reserve measures saving not as income minus consumption, but as additions to household financial net worth, such as household bank deposits, stock and bond portfolios, insurance policies, and the like. This approach permits us to measure net saving by the change in household net worth from year to year. (We make two adjustments: subtracting paper profits, that is, unrealized capital gains, from the value of stockholdings, and adding in assets omitted in Federal Reserve data, namely, Social Security net contributions and increases in the stock of owner-occupied housing.)
Figure 2 shows the savings rate from Federal Reserve Data along with the rates derived from the national income accounts. The main conclusion is clear. There has been no fall in personal saving. It is a statistical artifact, not an economic reality.
That saving did not fall should really be unsurprising, because the evidence was there all along, before our eyes. The purported fall in savings ill accords with trends of the last decade. As Kevin Phillips points out in The Politics of Rich and Poor, the top 5 percent of households, who account for a disproportionate share of saving, saw their real income increase by 23.4 percent from 1977 to 1988. The boom in stock market prices that began in the early 1980s hardly suggests a period of financial stringency. Neither does the past decade's massive wave of office-building indicate any shortage of credit. The downward trend of interest rates -- the return on a three-month Treasury bill dropped from 14 percent in 1981 to under 7 percent in 1988 -- does not agree with the conventional answer to the question: "What will happen to interest rates if saving dries up?"
The skeptical economist may object that rates would have risen had it not been for Japanese inflows of capital. Perhaps. But that simply changes the question from calculating the trend in saving to estimating its consequences. The availability of Japanese and other foreign capital suggests that even if there had been a fall in household saving, it would not have hurt investment because access to credit is today worldwide.
There remains one last, and perhaps most important, point. It concerns the serious misapprehension mentioned at the outset that only an increase in savings will revive the economy. This is a view of saving presumably banished by John Maynard Keynes when he published his epoch-marking General Theory of Employment, Interest, and Money in 1936.
Before Keynes, economists believed that savings were a hoard, like a cache of gold coins, which determined the ability of a nation to finance its capital-building operations. After Keynes, we have come to see saving in a different light -- not as a hoard, but as a flow that rises and falls with the level of economic activity. This flow of saving does limit our ability to create new capital when our economy is running at full blast, for we cannot then pry loose the labor needed for additional investment unless we can release it from its current employments. We do that by saving more, which means cutting back consumption to free up resources for investment projects. Hence, saving in a fully employed economy limits what we can invest, unless we are willing to unleash an inflationary contest among enterprises to attract, or retain, their labor.
But none of this, Keynes stressed, applied to an economy that was at less than "full" employment. Throughout the 1980s, we faced no such need to constrict consumption, because a fifth of our labor force was unemployed or only partially employed, and the same portion of our manufacturing capacity was standing idle. In underemployed times, the capacity to finance additions to plant and equipment is not limited by the release of otherwise fully occupied workers, but by the willingness and ability of banks to lend. That ability, in turn, is solely determined by the Federal Reserve Board of Governors.
Thus even if the saving rate had fallen, the lending capacity of the banking system or the expandability of the system would not have been significantly impaired. But that is not what the public has heard during recent years. Economists from both sides of the political fence have sounded the alarm in words that leave no room for uncertainty.
"It is widely recognized," said Lawrence Summers, then an economic adviser to Michael Dukakis, "that low national saving is the most serious problem facing the U.S. economy." Similarly, Alan Greenspan told us that "inadequate domestic savings is impairing our economic prospects for the long run. By choosing to consume more now and save less, we are limiting our ability to expand and upgrade our stock of capital." The New York Times, usually no friend of President Bush, editorialized, "Give him credit for focusing on the problem: Savings are needed for investment, and thus growth."
The lamentations would have been more to the point had they warned that our savings was as high as it was only because of capital gains and Wall Street shenanigans. But the economists and editorialists who predicted diminished growth from the disease of undersaving were saying, in effect, that our six to eight million unemployed were without jobs because households were spending too much. Recently, while wringing his hands over our failure to save enough, President Bush called on Americans to buy more cars and to take advantage of falling mortgage rates to acquire homes. This is more than a serious misapprehension. It is an absence of comprehension.
A capitalist economy in the doldrums can save as much as it wishes, but those savings, in and of themselves, will not generate the momentum that is wanting. In our view, the quickest, simplest, and most dependable way to revitalize a stagnating economy is to start up the engine of public investment, fueled with private savings. This not only uses saving for the very purpose that presumably justifies it -- financing growth -- but also creates more saving as the nation's households find their real income beginning to rise again. What is scarce in America is not saving, but political leadership and economic sense.