Chairman Ben S. Bernanke
Before The Washington Economic Club, Washington, D.C.
October 4, 2006
The Coming Demographic Transition: Will We Treat Future Generations Fairly?
In coming decades, many forces will shape our economy and oursociety, but in all likelihood no single factor will have as pervasivean effect as the aging of our population. In 2008, as the first membersof the baby-boom generation reach the minimum age for receiving SocialSecurity benefits, there will be about five working-age people (betweenthe ages of twenty and sixty-four) in the United States for each personaged sixty-five and older, and those sixty-five and older will make upabout 12 percent of the U.S. population. Those statistics are set tochange rapidly, at least relative to the speed with which one thinks ofdemographic changes as usually taking place. For example, according tothe intermediate projections of the Social Security Trustees, by2030--by which time most of the baby boomers will have retired--theratio of those of working age to those sixty-five and older will havefallen from five to about three. By that time, older Americans willconstitute about 19 percent of the U.S. population, a greater sharethan of the population of Florida today.
This coming demographic transition is the result both of thereduction in fertility that followed the post-World War II baby boomand of ongoing increases in life expectancy. Although demographersexpect U.S. fertility rates to remain close to current levels for theforeseeable future, life expectancy is projected to continue rising. Asa consequence, the anticipated increase in the share of the populationaged sixty-five or older is not simply the result of the retirement ofthe baby boomers; the "pig in a python" image often used to describethe effects of that generation on U.S. demographics is misleading.Instead, over the next few decades the U.S. population is expected tobecome progressively older and remain so, even as the baby-boomgeneration passes from the scene. As you may know, population aging isalso occurring in many other countries. Indeed, many of these countriesare further along than the United States in this process and havealready begun to experience more fully some of its social and economicimplications.
Even a practitioner of the dismal science like me would find itdifficult to describe increasing life expectancy as bad news. Longer,healthier lives will provide many benefits for individuals, families,and society as a whole. However, an aging population also creates someimportant economic challenges. For example, many observers have notedthe difficult choices that aging will create for fiscal policy makersin the years to come, and I will briefly note some of those budgetaryissues today. But the implications of demographic change can also beviewed from a broader economic perspective. As I will discuss, thebroader perspective shows clearly that adequate preparation for thecoming demographic transition may well involve significant adjustmentsin our patterns of consumption, work effort, and saving. Ultimately,the extent of these adjustments depends on how we choose--eitherexplicitly or implicitly--to distribute the economic burdens of theaging of our population across generations. Inherent in that choice arequestions of intergenerational equity and economic efficiency,questions that are difficult to answer definitively but arenevertheless among the most critical that we face as a nation.
Demographic Change and the Federal Budget
As I havealready mentioned, the coming demographic transition will have a majorimpact on the federal budget, beginning not so very far in the futureand continuing for many decades. Although demographic change willaffect many aspects of the government’s budget, the most dramaticeffects will be seen in the Social Security and Medicare programs,which provide income support and medical care for retirees and whichhave until now been funded largely on a pay-as-you-go basis. Undercurrent law, spending on these two programs alone will increase fromabout 7 percent of the U.S. gross domestic product (GDP) today toalmost 13 percent of GDP by 2030 and to more than 15 percent of thenation’s output by 2050. The outlook for Medicare is particularlysobering because it reflects not only an increasing number of retireesbut also the expectation that Medicare expenditures per beneficiarywill continue to rise faster than per capita GDP. For example, theMedicare trustees’ intermediate projections have Medicare spendinggrowing from about 3 percent of GDP today to about 9 percent in 2050--alarger share of national output than is currently devoted to SocialSecurity and Medicare together.
The fiscal consequences of these trends are large and unavoidable.As the population ages, the nation will have to choose among highertaxes, less non-entitlement spending, a reduction in outlays forentitlement programs, a sharply higher budget deficit, or somecombination thereof. To get a sense of the magnitudes involved, supposethat we tried to finance projected entitlement spending entirely byrevenue increases. In that case, the taxes collected by the federalgovernment would have to rise from about 18 percent of GDP today toabout 24 percent of GDP in 2030, an increase of one-third in the taxburden over the next twenty-five years, with more increases to follow.(This calculation ignores the possible effects of higher tax rates oneconomic activity, an issue to which I will return later.)Alternatively, financing the projected increase in entitlement spendingentirely by reducing outlays in other areas would require that spendingfor programs other than Medicare and Social Security be cut by abouthalf, relative to GDP, from its current value of 12 percent of GDPtoday to about 6 percent of GDP by 2030. In today’s terms, this actionwould be equivalent to a budget cut of approximately $700 billion innon-entitlement spending.
Besides tax increases, spending cuts, or reform of the majorentitlement programs, the fourth possible fiscal response to populationaging is to accommodate a portion of rising entitlement obligationsthrough increases in the federal budget deficit. The economic costs andrisks posed by large deficits have been frequently discussed and I willnot repeat those points today. Instead, I will only observe that, amongthe possible effects, increases in the deficit (and, as a result, inthe national debt) would shift the burden of paying for governmentspending from the present to the future. Consequently, the choices thatfiscal policy makers make with respect to these programs will be acrucial determinant of the way the economic burden of an agingpopulation is distributed between the current generation and thegenerations that will follow.
A Broader Economic and Generational Perspective
Indeed,framing the issue in generational terms highlights the fact that theeconomic implications of the coming demographic transition go wellbeyond standard considerations of fiscal policy and government finance,important as those are. For reasons that I will explain in a moment,the aging of the population is likely to lead to lower average livingstandards than those that would have been experienced without thisdemographic change. How that burden of lower living standards isdivided between the present and the future has important implicationsfor both intergenerational fairness and economic efficiency.
Why will the coming demographic transition carry a cost in terms oflong-run living standards? Assuming it unfolds as expected, theprojected aging of the population implies a decline over time in theshare of the overall population that is of working age and thus,presumably, in the share of the population that is employed. For anygiven level of output per worker that might be attained at some future date, this decline in the share of people working implies that the level of output per personmust be lower than it otherwise would have been. In a sense, eachworker’s output will have to be shared among more people. Thus, allelse being the same, the expected decline in labor force participationwill reduce per capita real GDP and thus per capita consumptionrelative to what they would have been without population aging. Thesereductions in output and consumption per person represent an economicburden created by the demographic transition.
Although some adverse effect of population aging on future percapita output and consumption is probably inevitable, actions that wetake today, in both the public and the private spheres, have thepotential to mitigate those effects. One such action would be to findways to increase our national saving rate. If the extra savings wereused to increase the nation’s capital stock--the quantity of plant andequipment available for use by workers--then future workers would bemore productive, ameliorating the anticipated effects on per capitaoutput and consumption. Alternatively, using extra saving to acquirefinancial assets abroad (or to reduce foreign obligations) would alsoincrease the resources available in the future.
By saving more today, we can reduce the future burden of demographicchange. However, as any economist will tell you, there is no such thingas a free lunch. Saving more requires that we consume less (to free upthe needed resources) or work more (to increase the amount of outputavailable to dedicate to such activities). Either case entails somesacrifice on the part of the current generation. Consequently, atradeoff exists: We can mitigate the adverse effect of the agingpopulation on future generations but only by foregoing consumption orleisure today. This analysis is simple, but it shows why the comingdemographic transition has economic implications that go well beyondthe effect of aging on the federal budget.
In recent work, economists at the Board of Governors have used astylized model to get a rough estimate of the magnitudes of theintergenerational tradeoffs that we face.1Their analysis takes as a starting point a baseline scenario in whichU.S. demographics remain (hypothetically) the same in the future asthey are today. In this counterfactual scenario, the ratio of workersto the overall population is assumed to remain at its current levelover time and per capita consumption grows with productivity. Now inreality, as I have noted, an aging population will reduce labor forceparticipation, so the likely future trajectory of per capitaconsumption over time lies below that implied by the baseline scenariothat assumes away the demographic change. The shape of the actualconsumption trajectory depends, however, on the saving behavior of thecurrent generation. If today’s saving rate is low, then the currentgeneration can enjoy consumption close to what it would have been ifthe aging issue did not exist. However, in this case, the burden onfuture generations will be relatively great. Alternatively, the currentgeneration could consume less and save more, which would allow theconsumption of future generations to be closer to what it would havebeen in the absence of population aging.
How big are these effects? To assess magnitudes, the Boardeconomists first examined the case in which the nation saves at itscurrent rate for the next twenty years, thereby largely insulating thebaby-boom generation from the effects of the coming demographictransition. After that, they assumed, consumption falls and savingrates rise, with all future generations experiencing the samepercentage reduction in consumption relative to the baseline in whichno population aging occurs. Their rough calculations suggest that, inthis case, the per capita consumption of future generations would beabout 14 percent less than what it would have been in the absence ofdemographic change.
For comparison, they next considered the case in which the burden ofdemographic change is shared more equally among current and futuregenerations. They considered a case in which the national saving rate,instead of staying at its current level for the next twenty years,rises immediately. Further, they asked by how much today’s saving ratewould have to increase to lead to equal burden-sharing among currentand future generations. ("Equal burden-sharing" is interpreted to meanthat the current generation and all future generations experience thesame percentage reduction in per capita consumption relative to thebaseline scenario without population aging.) They found that equalburden-sharing across generations could be achieved by an immediatereduction in per capita consumption on the order of 4 percent (or,since consumption is about two-thirds of output, by an increase innational saving of about 3 percentage points.) This case obviouslyinvolves greater sacrifice by the current generation, but the payoff isthat all future generations enjoy per capita consumption that is only 4percent, rather than 14 percent, below what it would have been in theabsence of population aging. The large improvement in the estimatedliving standards of future generations arises because of the extracapital bequeathed to them by virtue of the current generation’sassumed higher rate of saving.
These numbers shouldn’t be taken literally but the basic lesson issurely right--that the decisions that we make over the next few decadeswill matter greatly for the living standards of our children andgrandchildren. If we don’t begin soon to provide for the comingdemographic transition, the relative burden on future generations maybe significantly greater than it otherwise could have been.2
At the heart of the choices our elected representatives will have tomake regarding the distribution of these costs across generations willbe an issue of fairness: What responsibility do we, who are alivetoday, have to future generations? What will constitute ethical andfair treatment of those generations, who are not present today to speakfor themselves? If current trends continue, the typical U.S. workerwill be considerably more productive several decades from now. Thus,one might argue that letting future generations bear the burden ofpopulation aging is appropriate, as they will likely be richer than weare even taking that burden into account. On the other hand, I suspectthat many people would agree that a fair outcome should involve thecurrent generation shouldering at least some of that burden, especiallyin light of the sacrifices that previous generations made to give usthe prosperity we enjoy today.
The choice of which generations should bear the burden of populationaging has consequences for economic efficiency as well as forintergenerational equity. If we decide to pass the burden on to futuregenerations--that is, if we neither increase saving now nor reduce thebenefits to be paid in the future by Social Security and Medicare--thenthe children and grandchildren of the baby boomers are likely to facemuch higher tax rates. A large increase in tax rates would surely haveadverse effects on a wide range of economic incentives, including theincentives to work and save, which would hamper economic performance.Alternatively, to avoid large tax increases, the government coulddecide to sharply reduce non-entitlement spending in the future.However, such actions might also have important social costs that needto be taken into consideration.
Sharing the Burden of Population Aging
If, as a nation,we were to accept the premise that the baby-boom generation shouldshare at least some of the burden of population aging, what policysteps might be implied? As I have already noted, from a broad economicperspective, the most useful actions are likely to be those thatpromote national saving. Perhaps the most straightforward way to raisenational saving--although not a politically easy one--is to reduce thegovernment’s current and projected budget deficits. To the extent thatreduced government borrowing allows more private saving to be used forcapital formation or to acquire foreign assets, future U.S. output andincome will be enhanced and the future burdens associated withdemographic change will be smaller.
Increasing private saving, which is the saving of both the corporatesector and the household sector, is likewise desirable. Corporatesaving, in the form of retained earnings, is currently at relativelyhigh levels, but household saving rates are exceptionally low.3A broad-based increase in household saving would benefit both theeconomy and the millions of American families who currently hold verylittle wealth.
Unfortunately, many years of concentrated attention on this issue bypolicymakers and economists have failed to uncover a silver bullet forincreasing household saving. One promising area that deserves moreattention is financial education. The Federal Reserve has activelysupported such efforts, which may be useful in helping peopleunderstand the importance of saving and to learn about alternativesaving vehicles. Psychologists have also studied how the framing ofalternatives affects people’s saving decisions. For example, studiessuggest that employees are much more likely to participate in 401(k)retirement plans at work if they are enrolled automatically--with achoice to opt out-- rather than being required to actively choose tojoin. The pension bill recently passed by Congress and signed by thePresident included provisions to increase employers’ incentives toadopt such opt-out rules; it will be interesting to see whether suchrules are adopted and, if so, how effective they are in promotingemployee saving.
Other steps can also help increase the future productive capacity ofthe economy and thereby reduce the adverse effects of demographicchange. For example, devoting resources to improving our K-12 educationsystem, expanding access to community colleges, increasing on-the-jobtraining, and stimulating basic research could augment the nation’scapital in the broadest sense of the term and might have desirabledistributional effects as well.
Another response to population aging is to adopt measures thatencourage participation in the labor force, particularly among olderworkers. In the near term, increases in labor force participation wouldraise income; some of this income would be saved and would thus beavailable to augment the capital stock. In the long run, higher ratesof labor force participation, particularly by those who would otherwisebe in retirement, could help to offset the negative effect ofpopulation aging on the share of the population that is working.
To some extent, increased labor force participation by older workersmay happen naturally. Increased longevity and health will encouragegreater numbers of older people to remain longer in the workforce. Andslower growth in the labor force will motivate employers to retain orattract older workers--for example through higher wages, moreflexibility in work schedules, increased training directed toward olderworkers, and changes in the retirement incentives provided by pensionplans.
Reform of our unsustainable entitlement programs should also be apriority. The nature and timing of those reforms will be determined, ofcourse, by our elected representatives. However, the intergenerationalperspective does provide a few insights that might be helpful topolicymakers as they undertake the needed reforms. First, restructuringthe finances of our entitlement programs to minimize their reliance ondeficit spending will enhance national saving and reduce the burden onfuture generations. Second, changes in the structure of entitlementprograms should preserve or enhance the incentives to work and to save;for example, we should take care that benefits rules do not penalizethose who may wish to work part-time after retirement. Finally, theimperative to undertake reform earlier rather than later is great. Asillustrated by the simulation I discussed earlier, the longer the delayin putting our entitlement programs on a sound fiscal footing, theheavier the burden that will be passed on to future generations.Moreover, the sooner any restructuring of entitlement programs takesplace, the easier it will be for people now in their working years toprepare, for example, by saving more today. However, if reform isdelayed and fiscal exigencies ultimately force changes in theseprograms with little notice to potential retirees, their ability toadjust their behavior appropriately could be much reduced.
Conclusion
Over the next few decades, the U.S. populationwill grow significantly older, a development that will affect oursociety and our economy in many ways. In particular, the comingdemographic transition will create severe fiscal challenges, as thecost of entitlement programs rises sharply. I hope to have persuadedyou today, however, that the economic implications of this transitiongo well beyond fiscal policy. From a broader economic perspective, thequestion is how the burden of an aging population is to be sharedbetween our generation and the generations that will follow us. Afailure on our part to prepare for demographic change will havesubstantial adverse effects on the economic welfare of our children andgrandchildren and on the long-run productive potential of the U.S.economy.
References
Auerbach, Alan J., Jagadeesh Gokhale, and Laurence J. Kotlikoff(1991). "Generational Accounts: A Meaningful Alternative to DeficitAccounting," in David Bradford, ed., Tax Policy and the Economy, vol. 5, National Bureau of Economic Research. Cambridge, Mass.: MIT Press, pp. 55-110.
Elmendorf, Douglas, and Louise Sheiner (2000). "Should America Savefor Its Old Age? Fiscal Policy, Population Aging, and National Saving" Journal of Economic Perspectives, vol. 14 (Summer), pp. 57-74.
Gokhale, Jagadeesh, and Laurence J. Kotlikoff (2001). "Is War Between Generations Inevitable?" Report no. 246, National Center for Policy Analysis. Dallas, November.
Sheiner, Louise, Daniel Sichel, and Lawrence Slifman (2006). "APrimer on the Macroeconomic Consequences of Population Aging,"unpublished working paper, Board of Governors of the Federal ReserveSystem, Division of Research and Statistics, September.
Footnotes
1. See Sheiner, Sichel, and Slifman (2006) and Elmendorf and Sheiner (2000) for discussions of the basic approach. Return to text
2. Another approach for gauging the potentialimpact of demographic change on future generations is the generationalaccounting framework developed by Auerbach, Gokhale, and Kotlikoff(1992). This framework begins with the assumption that, for peopleliving today, tax rates will not be increased and benefits will not becut. On that assumption, one can calculate the taxes (net of transfersreceived) that future generations will have to pay to achieve long-termbalance in the government budget. According to recent estimates usingthis approach, to achieve long-term budget balance the net tax rate onfuture generations will have to be about double the tax rate on currenttaxpayers (Gokhale and Kotlikoff, 2001). This approach looks at theintergenerational issue through the prism of fiscal policy rather thantaking the broader economic perspective I have emphasized today, andits underlying assumptions are somewhat different. However, the basicmessage--that failure by the current generation to address the economicimplications of aging will impose significant costs on futuregenerations--is the same. Return to text
3. It is worth noting that a household’s savingneed not equal its change in wealth, since the standard definition ofsaving excludes capital gains. One plausible explanation of the recentlow level of household saving rates is that capital gains in stocks andin residential real estate, by increasing wealth, have reduced themotivation of households to save out of current income. If thatexplanation is correct, then the recent slowdown in the appreciation ofhouse prices should lead ultimately to some increase in householdsaving rates, all else equal. Return to text
