Printer Friendly

The saving solution.

The Savings Solution

IN RECENT YEARS it has become very popular among economists to blame many of the ills of the U.S. economy on a shortfall of savings. Thus, savings has often been the whipping boy for such intractable problems as productivity, investment, the trade balance, and the federal deficit.

Along with this emphasis on savings as a problem, the usual corollary is that something has to be done about it. As eighteen leading economists concluded at a Bank Credit Analysts' conference in Bermuda last spring: "The slowness in correcting the inadequacy of U.S. domestic savings was the single overriding concern about the U.S. economy and financial system."(1)


As Figure 1 shows, the savings rate has been much more stable since World War II than it was earlier. During the Depression the savings rate actually turned negative for three years, while during World War II it soared as high as 25 percent when incomes were high and consumer goods unavailable. What the chart doesn't show is the period prior to 1930. Between the Civil War and 1930 the U.S. savings rate was typically quite high and consistently exceeded the savings rates of Germany and Japan. In fact, over most of this period the U.S. had one of the highest savings rates in the world.

The earlier high-savings performance of the U.S. illustrates the important fact that savings rates are highest when a country is in its high-growth stage. When a country's economy is growing rapidly, high productivity gains allow both real earnings and savings to move up sharply. It is also at this stage of an economy's life cycle when it most needs investment to sustain growth. Thus, both savings and investment are high when countries are in the transition from a developing to a developed economy.(2) The other side of this coin is that mature economies require less in the way of savings and investment.

What bothers economists is that the savings rate was relatively stable in the postwar years until the mid-1970s and then started slipping. By 1987 the savings rate had reached its lowest level (3.2 percent) since 1947. The purpose of this analysis is to explain the reasons lying behind the secular decline of the savings rate and the implications for future savings trends in the U.S. economy.


To understand the meanings of savings trends, it is worthwhile to trace the evolution of savings theory. Prior to the 1930s little attention was paid to savings. But this changed as the world entered the era of the Great Depression. As an integral part of his theories, Keynes posited that a root cause of the depression was underconsumption, which, in turn, was due to excessive savings. A recent Public Agenda Foundation survey showed that this Keynesian view of savings being bad for the economy persists in the minds of the public.(3)

By the 1950s it became evident that savings theory required further refinement. In response, Franco Modigliani developed the life-cycle theory of savings that, in a nutshell, means people save for their old age. Thus, middle-aged persons envisioning retirement are high savers, while the young and the old tend to dissave. The life-cycle theory eventually also embraced the concepts that savings rise with income and fall with either increases in wealth or expected income.(4)

The life-cycle theory is the best explanation of savings behavior, but certain other theories also help explain the savings phenomenon. For example, Edward Denison found that the net private savings rate was relatively stable. The net private savings rate includes both personal and business savings. Thus, when household savings go up, business savings tend to go down. This observation makes sense because, when households save more, they obviously spend less. This cutback in consumer spending leads to declines in business profits and retained earnings, i.e., corporate savings. Conversely, when business profits rise, consumer wealth also goes up and individuals feel less need to save. Thus, the savings rate would then fall.(5)

Another theory has been offered by Michael Boskin, the Chairman of the President's Council of Economic Advisors. Boskin posited that savings were determined by both demographics and rates of return. In his view, savings respond to an increase in the after-tax rate of return,(6) implying that tax cuts raise savings. Thus, President Bush's proposal for a cut in capital gains taxes would be very beneficial to savings.

Others, however, have concluded that higher rates of return actually may discourage savings, because they make it easier for individuals to reach their savings goals. This situation appears to have been the case in the 1980s when higher interest rates have been coupled with low savings rates. Barry Bosworth has pointed out that pension fund trends provide an excellent example of the negative impact of higher interest rates upon savings, because overfunding often reduces or eliminates the need for contributions.(7)

Many other savings theories are afloat. But, most would, at best, provide only a very marginal explanation of recent savings trends. Included would be the idea that people save to pass on estates (Larry Summers), to pay off future federal deficits (Robert Barro), or because they are sanguine about the possibility of nuclear war (Jack Slemrod).


To determine which theory is most accurate, one has first to address the definition of savings. Many concepts of savings are calculated. Unfortunately, the one that draws the most attention is also the least accurate. This concept is that of the Department of Commerce savings rate shown in Figure 2. What is most troubling about this savings rate concept is that it is not calculated directly but is simply a residual. First, Disposable Personal Income is derived by subtracting taxes from personal income. Savings is then what is left after the combination of consumer spending, interest paid by consumers, and transfer payments to foreigners are subtracted from Disposable Personal Income.

The problems with this method of calculating a savings rate are many. For example, income can be severely underestimated due to the existence of the underground economy, while consumer spending estimates themselves can be very erroneous. This error partially results from large segments of consumer spending being imputed rather than calculated directly. These imputations can be especially distorting concerning consumer spending for housing, because the Commerce Department does not derive housing spending directly but instead uses what a house would rent for if available for rental.

The Federal Reserve Board also issues its own savings rate, which is little noticed but more accurate than the Commerce Department measure. The Federal Reserve's savings concept is based on information from financial institutions, monetary inflows, and estimates of the amount of cash held by households. It also adds in consumer durable goods purchases, government insurance credits, and capital gains. Its main fault is that it is often distorted by international investment flows.

The Federal Reserve savings measure, as well as the Commerce Department savings rate, have both trended downward since the mid-1970s. But the Federal Reserve savings rate is still more than 2 1/2 times as high as the Commerce Department rate. I should also note that the relative positions of the two savings rates have reversed over time. Back in the 1950s the Commerce Department savings rate was 4 percentage points higher than the Federal Reserve rate. By 1988, however, the Commerce Department rate was 6.2 percentage points lower.(8)

Because the Commerce Department savings rate is the one used as the basis for most comparisons, it is worthwhile to point out its limitations. Specifically, the Commerce Department savings rate excludes Social Security, capital gains, Federal retirement plans, education and training, home equity build-up and increases in life insurance cash values. Also excluded are investments in consumer durables (Hendershott and Peak estimate that adjusting for consumer durables alone would add 3 percentage points to the official savings rate)(9) as well as merger and acquisition payouts (last year's mergers and acquisitions created a flow of income equal to 4 percent of Disposable Personal Income from the corporate to the household sector).(10)


Many forms of savings are available to individuals, most of which should be covered by the official savings concept. Table 1 shows the increases in various forms of savings between 1975 and 1988. Over this period net savings, as portrayed by the Commerce Department, rose 151 percent. Yet all forms of savings shown climbed faster with the singular exception of life insurance equity. On the top are money market funds which surged 8,014 percent. Even household net worth, which includes capital gains, rose much faster than savings. This listing makes it evident that some real problems exist in using the savings rate as a guide to actual savings trends.


(Percent Changes)
Money Market Funds 8,014%
Government Securities Funds 422
Pension Funds 414
Land Holdings 387
Corporate Equity Holdings 250
Stock of Consumer Durables 208

Checkable Deposit Accounts
and Currency 204
Time and Savings Accounts 195
Life Insurance Equity 88
Net Savings 151
Household Net Worth 214


Why has the savings rate declined? Many explanations have been proffered, such as demographic changes, slower income growth, reduced income volatility (due to the increasing numbers of two-earner families), rising wealth, disinflation (expectations of higher prices induce consumers to save not spend), and the strong housing market (purchases of housing cut the savings rate because mortgage debt reduces savings while the equity buildup is not incorporated in the savings concept).

Two factors in particular have been credited with having a significant negative impact on the savings rate. One is cuts in pension contributions due to overfunding, which has been estimated to have had a full percentage point negative impact on the savings rate. The other is the better financial position of the elderly, which may have convinced many younger persons that they do not have to save a lot for retirement.(11) Between 1981 and 1987, the median income for persons 65 years and older grew far faster than for any other age group. Another factor is that the availability of Social Security alone probably knocked the savings rate down by a percentage point.(12)

In addition, individuals may now feel less need to save for emergencies, because more of their hospital expenses are now covered by insurance. In 1950, 30 percent of hospitalization payments were made by individuals compared to 9 percent in the mid-1980s.(13) Improved disability and life insurance plans also possibly could have reduced savings needs.

Despite all of these factors favoring slowing savings, the National Bureau of Economic Research has concluded that almost all of the decline in savings during the 1980s would have been wiped out if adjustments were made for such factors as consumer durables investment and federal retirement plans.(14)

What savings rate should we use? Table 2 covers four different calculations of the savings rate. The first column shows the 1988 rate, the second column the average savings rates for the "high savings" 1970-75 period. The first concept, "personal savings" is the Commerce Department measure and it indicates a definite falloff over this period. The "flow of funds" concept is the Federal Reserve measure, which also illustrates a decline but nowhere near as steep.

Table : Table 2

What Are The Different Concepts Of Savings?
 1988 1970-75
 Rate Rate
Personal Savings 4.2% 8.6%
Flow of Funds 10.4 13.1

(Adds Depreciation of Consumer Durables, Capital Gains, etc.)

Gross Private Savings(*) 15.0 17.4 (Includes Depreciation)

Gross Savings(*) 13.2 16.3 (Includes Government)

(*)Percent of GNP

The third concept "gross private savings" includes business savings as well as consumer savings and has been very stable over the past thirty-five years. As Paul Craig Roberts has pointed out, the rise in business savings during the 1980s (largely due to faster write-offs) has offset the decline in personal savings.(15) The final concept, "gross savings," is the broadest of all and includes government savings. This concept also has been very stable for thirty-five years.


The point of the table is that what is really important is the total amount of savings available for investment in this country. Virtually all of the arguments regarding an imbalance between savings and investment in this country utilize the Commerce Department savings rate to back their arguments. Yet, the amount of funds available for investment is determined by total savings in the U.S., not just that generated by consumers. Thus, the stable broader concepts of "gross private savings" and "gross savings" are far better guides to the total funds available for investment than the very narrowly defined "official" savings rate.

Also, the links between savings and investment as well as between investment and productivity are tenuous at best. For example, the mere provision of savings does not necessarily imply more investment, because such savings can be used for speculation or investment in another country.(16) And, as Lou Ferleger and Jay Mandle have pointed out, the investment rate for 1983 through 1987 exceeded that for 1961-68. Yet productivity growth was significantly higher in the earlier period.(17)


Savings rate arguments also revolve around the U.S. savings rate being far below those of other countries when adjusted for differences in accounting methodologies.

If these differences actually exist, there are some sound explanations for them:

1. The other developed countries have fewer people than the U.S. in the high-borrowing age groups. Consumer borrowing reduces the savings rate by definition and, even for those in the high-borrowing age groups, less credit is available in other countries.

2. The U.S. fully taxes savings while other countries typically do not tax savings or, even if they do, have a harder time collecting the taxes. For example, Germany allows tax free bonuses; the United Kingdom has tax-free retirement bonds; Canada defers taxes on employer contributions to pension plans; the Japanese Postal Savings System has, in the past, been a huge repository of tax-free savings.

3. The Social Security systems in other countries are less protective than in the U.S. In Japan, for example, Social Security starts at 60 but lifetime jobs end at 55, leaving five years of retirement to be financed solely out of savings.

4. Homes are bought later in life in other countries and down payments are far larger. The Japanese typically are in their forties before they can buy their first house, the down payment is usually around 40 percent. Within 1 1/2-2 hours of Tokyo, the home value is typically nine times annual income (compared with 3.3 times in the U.S.). In addition, mortgage interest is not deductible.(18)

5. In most other countries, a larger share of consumer incomes comes in the form of year-end bonuses. Such lump sums are much more likely to be saved than true for other income.

6. In terms of investment needs, the U.S. does not require as high a savings rate as other countries, because structures and equipment are typically lower priced in the U.S. (a factor that would alone reduce the savings rate differential by several points).

It is commonly thought that U.S. investment has declined over time. It has not. As a share of GNP, U.S. investment has been stable. The difference is that the investment share of GNP has been accelerating in other countries. Many reasons explain this relationship, none of which imply that the U.S. has not been pulling its weight. Among these reasons are the spread of technology from the U.S. to the rest of the world and a later shift out of agriculture in other countries.

In the end, though, faster investment growth in other countries is a catch-up phenomenon. What is most important is that the U.S. capital stock per capita is still the highest in the world.

All of these points support the view that persons in other countries, by necessity, have to save more than those in the U.S. But, in spite of the validity of this argument, still some question remains as to whether they are actually saving more. In an intriguing analysis, the managing director of McKinsey & Co. in Japan, Kenichi Ohmae, has negated most of the differences between the U.S. and Japanese savings rates by adjusting for international calculation standards, accounting inconsistencies, and socioeconomic differences (such as higher home prices and down payments in Japan). The result is that the adjusted Japanese savings rate only exceeds the U.S. rate by a small amount.

In a further refinement of his argument, Ohmae makes the very valid point that the real issue is not the flow of savings as represented by the savings rate but the actual outstanding stock of savings. On this basis, U.S. per capita savings stocks are three times higher than those of Japan.(19) This difference is due to the accumulation of U.S. savings over time and higher interest rates in the U.S. He then carries his analysis one step further by dividing increases in household assets by disposable income for both countries. When this calculation is made, the Japanese savings rate of 14.4 percent in the 1980s is actually below the 19.6 percent of the U.S.(20)


We looked at all of the factors thought to influence savings, many of which have good graphic relationships to the U.S. savings rate. Included among these factors were income, wealth, interest rates, spending on energy, inflation, credit, and the trade balance. But a regression analysis showed that only two variables have a significant relationship to the U.S. savings rate. One is existing home sales, where the relationship was marginally significant. The other was demographics, especially the share of the population in the high-saving 45 to 69 year age group. Demographics proved to be a highly significant predictor of the savings rate.

Figure 3 shows the close relationship between the 45 to 69 year old population share and the savings rate while Figure 4 illustrates the importance of the coming demographic changes. Starting in 1993, the 45 to 69 year-old segment of the population will start soaring as a share of the total. At the same time the 25 to 44 population group's share will decline sharply. The major demographic impact of these changes will occur between 1995 and 2010.

The results of these demographic changes on the savings rate are shown in Figure 5. We believe that in 1993 the savings rate will turn around and then accelerate upward. By 2010 the savings rate will be challenging the peaks of the early 1970s.


There are two camps of economists regarding the significance of future savings trends. Some believe that demographics will have a very definite impact upon savings in the future, while others believe that the effects will be marginal.

Among the latter, the argument is often given that the income shares of different age groups are relatively stable over time. Thus, a smaller 45 to 64 population cohort, for example, would be largely offset by higher wages due to increased demand for workers in this age group. Table 3 shows, however, that income shares are positively related to population changes.


Age Group 1972-73 1986 Change

25 to 44 Years Old
 Share of Consumer Units 36.8% 42.5% +5.7%
 Share of Income 41.4 49.5 +8.1

45 to 64 Years Old
 Share of Consumer Units 34.3 27.5 -6.8
 Share of Income 41.5 33.1 -8.4

65 Years and Over
 Share of Consumer Units 20.0 20.5 +0.5
 Share of Income 11.9 12.8 +0.9

Others base their conclusions on the small variations in savings rates by age group often found in consumer surveys. But the problem with such surveys is that it is very difficult for interviewers to elicit accurate information on such concepts as savings, wealth, and income from respondents. For one thing, individuals have a hard time accurately estimating savings flows and increases in wealth.

In addition, a high proportion of respondents refuse to answer questions on such sensitive concepts, and this refusal rate is much higher for those with substantial assets. Even those wealthier people who do answer the questions would tend to underestimate the dollars involved significantly. The result is not only extremelypoor estimates of savings rates by age but also sharp shifts in such rates from survey to survey.

For example, the age group normally thought to have the highest savings rate would be the 45 to 54 year olds. Yet the BLS Consumer Expenditures Survey showed a negative savings rate for this age group in 1986.

Another argument for continuing low savings rate has been advanced by economists at Salomon Brothers(21) and Goldman Sachs.(22) They have posited that the future higher savings of the 45 to 59 year olds will be offset by dissaving among persons 60 years and over. They base their findings on median savings ratios by age from the 1983 and 1986 Surveys of Consumer Finances. The problem with this approach is that the overall savings rate is determined by mean savings ratios rather than medians. As the Federal Reserve has concluded, the skewness of wealth is strongly related to savings. This factor explains why the Survey of Consumer Finances shows that the median savings ratio of 60 to 74 year olds is lower than that for any age group under 60 years while their mean savings ratio is higher than those of all other age groups.(23) A basic rule of statistics is that you do not multiply a median by the number in a group to get an accurate total.


Thus, we believe that demographic changes are the prime determinants of the U.S. savings rate and, over the next twenty years, these changes will push the savings rate up sharply. As John Rutledge has aptly said, "Rising savings rates will make the U.S. the Japan of the 1990s with very low inflation and super-low interest rates, rising capital formation, growing productivity and an improving trade situation."(24)

In fact, ten years from now the debate is likely to concern savings rates that are too high rather than too low. Some have proposed that a value-added tax be implemented to stimulate savings. A value-added tax is the worst possible way to try to boost savings. Such a tax would lower spendable income and force consumers to maintain their spending levels by cutting their savings. Thus, a value-added tax would reduce, not increase savings.

As Alan Greenspan has said: "A sharp contraction in the federal deficit appears to be the only assured source of augmenting domestic net savings." In his testimony before the National Economic Commission, Greenspan stressed that this reduction in the deficit should come from spending cuts, not revenue increases. His reason: the proclivity of Congress to use additional revenues generated by tax hikes to increase spending. We could not agree more.


(1)Boeckh, Dr. J. Anthony, Bank Credit Analyst, Synopsis of May 16-17, 1988 Conference.

(2)Modigliani, Franco, "The Key To Saving is Growth Not Thrift," Challenge, May/June, 1987.

(3)"Saving: Good or Bad," The Public Agenda Foundation, 1989.

(4)Modigliani, Franco, "Life Cycle, Individual Thrift and the Wealth of Nations," American Economic Review, June, 1986.

(5)Denison, Edward F., "A Note On Private Savings," Review of Economies and Statistics, August, 1958.

(6)"Reaganomics Effect on Savings," Business Week, March 8, 1982.

(7)Bosworth, Barry, Testimony before House Ways and Means Committee, April 19, 1989.

(8)Curtin, Richard, "Consumer Saving and Spending Prospects for the Decade Ahead," University of Michigan, May, 1982.

(9)Hendershott, Patrick and Joe Peek, Forbes, September 7, 1987.

(10)Summers, Laurence and Chris Carroll, "Why is U.S. National Savings So Low," Brookings Papers on Economic Activity, February, 1987.


(12)Sylvester, David, "Is the Saving Rate Really That Bad," Fortune, November 7, 1988.

(13)Summers, Laurence, op. cit.

(14)Morgan, John T., "Savings in the U.S. Dead?" Bottom Line, March, 1988.

(15)Roberts, Paul Craig, "Why America's Piggy Banks Aren't Bulging," Business Week, June 20, 1988.

(16)Tyler, Gus, "To Spend or To Save," book review in Challenge, March/April, 1989.

(17)Ferleger, Lou and Jay R. Mandle, "The Savings Shortfall," Challenge, March/April, 1989.

(18)Collins, Catherine, "Chasing The Japanese Dream," Secondary Mortgage Markets, Fall, 1988.

(19)Ohmae, Kenichi, "Americans and Japanese Save About the Same," Wall Street Journal, June 14, 1988.

(20)Ohmae, Kenichi, "Is Japanese Savings Rate Too High," Japan Economic Journal, May 14, 1988.

(21)Berner, Richard, "Don't Count on a Yuppie Savings Boom," Salomon Brothers Bond Market Research Memorandum, August 10, 1988.

(22)Benderley, Jason and Edward McKelvey, "The U.S. Saving Shortfall: No Easy Way Out," Goldman Sachs, The Pocket Chartroom, October/November, 1988.

(23)Avery, Robert B. and Arthur B. Kennickell, "Savings and Wealth: Evidence from the 1986 Survey of Consumer Finances," presented at Conference on Research in Income and Wealth, May 12-14, 1988. See Chart 3.

(24)Brimelow, Peter, "The Saving of America," Forbes, April 20, 1987.

PHOTO : Figure 1 Since World War II The Savings Rate Has Averaged 6%

PHOTO : Figure 2 The Two Savings Rates

PHOTO : Figure 3 A Higher Proportion Of 45 To 69 Year Olds Tends To Keep The Savings Rate High

PHOTO : Figure 4 Demographics Will Play A Key Role In Future Savings

PHOTO : Figure 5 Demographics Should Drive The Savings Rate Up
COPYRIGHT 1989 The National Association for Business Economists
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1989 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Swanstrom, Thomas E.
Publication:Business Economics
Date:Jul 1, 1989
Previous Article:Small Island Economies: Structure and Performance in the English-Speaking Caribbean Since 1970.
Next Article:Analyzing consumer spending and debt.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters