1、October 1, 1990eCONOMICCOMMeNTORYFederal Reserve Bank of ClevelandDont Worry, Well Grow Out of It:An Analysis of Demographics,Consumer Spending, and Foreign Debtby Michael F. Bryan andSusan M. ByrneM,distrustful might be a moreexplanatory word than authoritarian,I think. Parents with this slant tend
2、 tobe stern because they assume thatchildren, if left to themselves, will bemore inclined to be naughty thangood Dr. Benjamin SpockOver the past 20 years, the view hasemerged that Americans are spendingtoo much and saving too little, financ-ing this spending binge, in large part,with foreign investm
3、ent. In this view,we have mortgaged future increases inour standard of living to foreigners inreturn for a spendthrift lifestyle today.Concerned policymakers have advo-cated several strategies designed toforce us to save more, such as consump-tion, gasoline, and import taxes, andother measures inten
4、ded to discourageconsumer spending. But before adopt-ing any drastic proposals, we shouldconsider carefully whether consumerspending has, in fact, been excessive.This Economic Commentary focuseson demographic changes in the UnitedStates since the late 1960s and theirimplications for consumer spendin
5、g,debt, and foreign investment. Using thelife-cycle theory of consumption as aguide, we have calculated a simple,forward-looking measure of the lifetimeearnings potential of the labor force forthe years 1968 through 1989. Ourestimates show that the inrush of thebaby-boom generation into theworkforce
6、 during the 1970s and 1980sproduced a substantial increase in theaggregate lifetime earnings potential ofU.S. households relative to their cur-rent labor income, which may help toexplain the otherwise worrisome spend-ing and debt patterns that developedover this period.Furthermore, as the baby-boome
7、rs reachmiddle age, we should see a reversal inlong-term spending and debt trendsaturnabout that our estimates suggestmay already be underway. Well Gladly Pay You Tuesday-Consumer spending as a share ofnational income was higher in the1980s than at any other time during thelast century. In 1987, for
8、 example, con-sumption claimed 82 percent of U.S.national incomeup from 75 percent inthe late 1960spresumably the highestspending-to-income ratio of any majorindustrialized nation. The flip side, ofcourse, is the personal savings shortfall:The U.S. personal savings rate fell fromaround 7 percent of
9、national income in1970 to an average of 4/2 percent in thepast five years.While consumers were boldly spendingmore and saving less, so, too, was thefederal government. Between 1970 and itsBetween the mid-1960s and early1980s, the age distribution of the U.S.labor force was changed dramaticallyby the
10、 inrush of the baby-boom gener-ation. The authors examine the impli-cations of this shift for consumerspending, debt, and foreign investment,and conclude that, if left to themselves,consumers will simply outgrow theirapparently spendthrift ways.peak in 1986, the federal deficit rosefrom $12.4 billio
11、n to $207 billion, orfrom 1 Vi percent to 6 percent ofnational income. Although the savingrates of businesses and state and localgovernments actually improved betweenthose same yearsfrom about 13 per-cent to 173/4 percentthe increaseswere insufficient to offset the drop-offthat occurred in household
12、 and federalgovernment saving rates. Over this 16-year period, national saving (includinghouseholds, businesses, and the govern-ment) dropped from 18/2 percent ofnational income to 15/2 percent.This decline coincided with a rise inU.S. debt accumulation, from 17 per-cent to 24!/2 percent of national
13、income between 1970 and 1986. Therun-up in federal government debt overthis period was dramatic, but house-ISSN 0428-1276FIGURE 1 THE AGE/EARNINGS PROFILE: MEDIANAVERAGE WEEKLY EARNINGS, 1968-88Index, 20-year-olds = 1002252001-11111111111111111111 lii 111111111111100 -7520 60SOURCE: Current Populati
14、on Survey, U.S. Department of Commerce, Bureau of the Census, 1968 through1988.FIGURE 2 PERSONAL SAVINGS AND NETFOREIGN INVESTMENT“1970 1975 1980 1985 1989a. As shares of national income.SOURCE: U.S. Department of Commerce, Bureau of Economic Analysis.hold liabilities, particularly consumerinstallme
15、nt credit, also ballooned. Theratio of consumer installment debt tonational income rose from about 121/2percent in the early 1970s to 163/t per-cent by 1986.It follows that an individual who borrowsmore is probably also one who saves less.Yet, how is it that a nation can borrowmore and at the same t
16、ime save less?Where do the funds come from for bor-rowers to borrow? Enter foreign savers.In the early 1980s, the United Statesran a large and growing trade deficit,the result of which was a huge inflowof foreign capital. But ultimately, alldebts, foreign or otherwise, must berepaid, and therein lie
17、s the problem.High levels of consumer spending and acorresponding reduction in personalsaving suggest an investment trend thatis insufficient to satisfy these debts.The long-term repercussions of foreignindebtedness depend, in part, on howthese borrowed funds are used. Pre-vious analysis by this ban
18、k indicatesthat roughly half of the inflow offoreign capital during the currentexpansion financed higher levels ofinvestmentnamely, business spend-ing on physical capital and consumerspending on durable goods and educa-tion. These expenditures bode well forfuture U.S. living standards.Nevertheless,
19、a significant share offoreign investment seems to havefinanced what we consider to be “pure“consumption, that is, consumer spend-ing on nondurable goods and services. . for a Hamburger Today.A common interpretation of U.S. spend-ing and debt trends is that a multitude ofAmericans are being naughty i
20、n theirspending habits, the upshot of whichwill be financial hardship at some futuredate. In this scenario, spendthrift con-sumers do not fully appreciate the long-term implications of their actions. Ifthey did, they would surely act in amore fiscally responsible way.We prefer an alternative explana
21、tionone that assumes that individualsactions are based on a reasonable under-standing of their current and expectedfuture financial condition. This explana-tion employs the life-cycle theory ofconsumer spending, where consump-tion is not bridled by current income butis instead based on an individual
22、sexpected lifetime earnings.Consider the average U.S. life cycle.Workers by and large enter the labormarket between the ages of 18 and 22and retire from regular employment intheir mid-60s, over which time theiraverage earnings vary dramatically (fig-ure 1). Weekly earnings soar during thefirst 10 ye
23、ars of an individuals worklife, such that workers in their mid-30searn roughly twice as much per week asthose in their early 20s. Weekly earn-ings remain relatively constant untilworkers enter their 60s, at which timetheir attachment to the labor marketweakens and they increasingly rely onaccumulate
24、d wealth as a source ofincome. A commonly accepted expla-FIGURE 3 AGE DISTRIBUTION OF WORKERS, 1965 vs. 19801965 1980Percent PercentSOURCE: U.S. Department of Labor, Bureau of Labor Statistics.nation for the age/earnings profiledepicted in figure 1 is that the experi-ence workers gain during the fir
25、st 10years of their working lives substan-tially increases their productivity and,hence, their income.We see, then, that a workers lifetimeearnings potential, or life-cycle earn-ings, depends crucially upon age.Young workers have a large lifetimeearnings potential relative to their cur-rent earnings
26、first, because they havea long expected worklife, and second,because they expect substantial wageincreases as their experience in theworkplace grows. However, spendingpatterns do not typically follow theearnings stream, as consumers attemptto maintain a more even standard ofliving relative to their
27、income overtime. They do this by borrowing againsttheir expected future earnings from per-sons who are in a later stage of the lifecycle.Because of these different patterns ofincome and spending over the course ofa life cycle, younger workers tend tohave high levels of spending and debtrelative to t
28、heir current income, whilemiddle-aged workers are inclinedtoward relatively low levels of spendingand indebtedness. An extreme shift inthe composition of the labor forcetoward young workers should induce anincrease in a nations desired consump-tion and debt relative to its income, asthe younger work
29、ers attempt to borrowand spend against their expected futureearnings.As long as a balance exists betweenyouthful borrowers and middle-agedsavers, harmony in national creditmarkets will prevail. But if the level ofdesired borrowing exceeds the level ofdesired saving (which should haveoccurred when th
30、e baby-boom genera-tion entered the workforce), upwardpressure on interest rates will result.Without an external source of funds,the rise in interest rates would serve toration relatively scarce credit; someindividuals would be prevented fromachieving their preferred lifetime con-sumption path, and
31、some would “crowdout“ other borrowersnamely, busi-ness investors. However, in a globalsetting, the rise in domestic interestrates relative to foreign rates producedby a credit shortfall encourages aninflow of investment from abroad, thusallowing a nation to simultaneouslysave less and borrow more. I
32、ndeed, thetrends in personal savings and net for-eign investment have been strikinglysimilar since 1973 (figure 2).Therefore, a major shift in the age com-position of the labor force (not occur-ring simultaneously in other countries)can produce changes in internationaldebt and trade flows, as nation
33、s withrelatively younger workers borrowfrom nations with relatively olderworkers. Consequently, long-term con-sumer spending and debt patterns canbe misleading if the age distribution ofthe labor force is not considered. Magical Mystery TourOver the 15-year period between 1965and 1980, the age distr
34、ibution of theU.S. labor force shifted dramatically asthe baby-boomers charged into the labormarket (figure 3). In 1965, about 34 per-cent of all adult workers were under theage of 34, compared with almost 47 per-cent by 1980. Meanwhile, the share ofthe middle-aged workforce (between 35years and 59
35、years) shrank from about60 percent to 49 percent.To measure the influence of the baby-boom generation on the growth of life-time earnings potential relative to cur-rent income, we constructed an estimateof the lifetime income potential of thelabor force using annual survey data forthe years 1968 thr
36、ough 1989.7 Thelifetime earnings potential of anemployed 24-year-old is estimated as thecurrent earnings of a 24-year-old, plusthe present value of the current earningsof a 25-year-old adjusted for one year ofproductivity growth, plus the presentvalue of the current earnings of a 26-year-old adjuste
37、d for two years ofproductivity growth, and so on, until thatworker retires at an assumed age of 65years. To find the lifetime earningspotential of the labor force in any par-ticular year, we multipy the lifetime earn-ings potential of each age group in thatyear by the number of workers currentlyin t
38、hat age group, and then sum over allage groups.We make several simplifying assump-tions in the construction of these esti-imates. For example, workers who aretemporarily out of the labor market dur-ing the annual survey are not included inour lifetime earnings calculations. Con-sequently, the data a
39、re more sensitive tothe national business cycle than theyshould be. We also assume that workerscurrently in the labor force expect toremain there, without interruption, untilage 65. That is, we neither allow forperiods of unemployment (planned orunplanned) nor for the possibility ofdeath prior to th
40、e age of retirement.Further, our methodology makes noadjustment for taxation. We acknow-ledge that permanent changes in eitherincome taxation or social security pay-ments that are perceived as permanenttaxation could lower the lifetime earn-ings estimate as it applies to pure con-sumption. However,
41、despite these simpli-fications, we believe that this seriescaptures the broad impact of a majorinflux of young workers on the potentialearnings stream of the labor force.Figure 4 shows the lifetime earningspotential of the labor force relative toits current income for the years 1968FIGURE 4 LIFETIME
42、 EARNINGS POTENTIAL ANDNONDURABLES AND SERVICES CONSUMPTION Lifetime earnings potential1970 1975 1980 1985 1989a. As a share of current income.b. As a share of national income.SOURCES: U.S. Department of Commerce, Bureau of Economic Analysis; and the Federal Reserve Bank ofCleveland.RatioFIGURE 5 LI
43、FETIME EARNINGS POTENTIAL3iSU29282726251 I I 1 I 1 1 I 1 1 1 1 1 1 1 1 1 1 1 1 1 1.1 1 1 1 1 1 1 1 1 11970 1975 1980 1985 1990 1995 2000a. As a share of current income.SOURCE: The Federal Reserve Bank of Cleveland.through 1989.10 Potential lifetime earn-ings rose at a rate of 1 lA percent moreper ye
44、ar than current income between1968 and 1984, and the ratio of poten-tial lifetime earnings to current incomerose from about 24/2 to 29. That is, byvirtue of its relatively youthful stand-ing, the labor force in 1984 expectedan additional 18 percentor 4Vi yearsof earnings relative to that expectedby
45、the labor force in 1968. Potentiallifetime earnings peaked in the mid-1980s and, according to our estimates,began to decline relative to currentincome in 1986.Trends in the ratio of lifetime earningspotential to current earnings correspondroughly with trends in the share ofnational income spent on c
46、onsumer non-durable goods and services, or “pure“consumption (figure 4). Indeed, pureconsumer spending as a share of nationalincome peaked at roughly 71 Vi percentin 1983, and since 1987, this percentagehas been declining. They Grow Up So Fast, DontThey?In 1986, the oldest members of thebaby-boom ge
47、neration celebrated their40th birthdays. Since then, the lifetimeearnings potential of the labor forcehas fallen abruptly relative to currentincome. This reflects a labor forcenear the peak of its age/earningsprofile, with an average age that hasbeen rising. Consistent with thesetrends, the rate of
48、personal saving hasescalated, and we have begun to reduceour dependence on foreign capital.The U.S. labor force is expected tobegin aging rapidly. According to theU.S. Census Bureau, the 20- to 34-year-old age group is projected to fall fromabout 43 percent of the working-agepopulation to around 35
49、percent overthe next 10 years. As a result, weproject that the ratio of lifetime earningspotential to current income will dipfrom 28 percent to less than 26 percentbetween the years 1990 and 2000 (fig-ure 5). This trend will likely exertupward pressure on personal savingrates and downward pressure on realU.S. interest rates, discouraging foreigncapital inflows in the process.We believe that a compelling case canbe made for the importance of agedemographics on the determination ofthe aggregate consumption appetite ofU.S. households. We recognize, how-ever, that se
