《麥肯錫：2025生育率下降與長壽時代雙輪驅動下的人口結構范式遷移：面對人口新現實的影響研究報告（英文版）（82頁）.pdf》由會員分享，可在線閱讀，更多相關《麥肯錫：2025生育率下降與長壽時代雙輪驅動下的人口結構范式遷移：面對人口新現實的影響研究報告（英文版）（82頁）.pdf（82頁珍藏版）》請在三個皮匠報告上搜索。

1、Confidential and proprietary.Any use of this material without specific permission of McKinsey&Company is strictly prohibited.January 2025Authors Anu MadgavkarMarc Canal NoguerChris BradleyOlivia WhiteSven SmitTJ RadiganEditorStephanie StromData visualizationJuan M.VelascoDependency and depopulation?

2、Confronting the consequences of a new demographic realityCopyright 2024 McKinsey&Company.All rights reserved.Cover image:A silhouetted child soars through the air on a swing set.The setting sun casts a warm glow on the scene,creating a quiet and nostalgic atmosphere.SchulteProductions/Getty ImagesCo

3、nfidential and proprietary.Any use of this material without specific permission of McKinsey&Company is strictly prohibited.McKinsey Global InstituteThe McKinsey Global Institute was established in 1990.Our mission is to provide a fact base to aid decision making on the economic and business issues m

4、ost critical to the worlds companies and policy leaders.We benefit from the full range of McKinseys regional,sectoral,and functional knowledge,skills,and expertise,but editorial direction and decisions are solely the responsibility of MGI directors and partners.Our research is currently grouped into

5、 five major themes:Productivity and prosperity:Creating and harnessing the worlds assets most productively Resources of the world:Building,powering,and feeding the world sustainably Human potential:Maximizing and achieving the potential of human talent Global connections:Exploring how flows of goods

6、,services,people,capital,and ideas shapeeconomies Technologies and markets of the future:Discussing the next big arenas of value and competitionWe aim for independent and fact-based research.None of our work is commissioned or funded by any business,government,or other institution;we share our resul

7、ts publicly free of charge;and we are entirely funded by the partners of McKinsey.While we engage multiple distinguished external advisers to contribute to our work,the analyses presented in our publications are MGIs alone,and any errors are our own.You can find out more about MGI and our research a

8、t Directors Sven Smit(chair)Chris Bradley Kweilin Ellingrud Sylvain JohanssonNick LeungOlivia White Lareina YeeMGI Partners Mekala Krishnan Anu Madgavkar Jan Mischke Jeongmin Seong 1Dependency and depopulation?ContentsAt a glance 3Introduction 41.The age of youth scarcity 62.A lower-growth,higher-de

9、pendency future in first wave regions 242.1.Demographic shifts,slower economic growth 252.2.More workers,more hours,and more productivity needed 352.3.A growing societal bill for dependency 442.4.Older workers,older consumers 553.Anticipating demographic shifts in later wave regions 624.Responding t

10、o a new demographic reality 71Acknowledgments 75Glossary 76Endnotes 772Dependency and depopulation?At a glance Falling fertility rates are propelling major economies toward population collapse in this century.Two-thirds of humanity lives in countries with fertility below the replacement rate of 2.1c

11、hildren per family.By 2100,populations in some major economies will fall by 20 to 50 percent,based on UN projections.Age structures are invertingfrom pyramids to obelisksas the number of older people grows and the number of younger people shrinks.The first wave of this demographic shift is hitting a

12、dvanced economies and China,where the share of people of working age will fall to 59percent in 2050,from 67 percent today.Later waves will engulf younger regions within one or two generations.Sub-Saharan Africa is the only exception.Consumers and workers will be older and increasingly in the develop

13、ing world.Seniors will account for one-quarter of global consumption by 2050,double their share in 1997.Developing countries will provide a growing share of global labor supply and of consumption,making their productivity and prosperity vital for global growth.The current calculus of economies canno

14、t support existing income and retirement normssomething must give.In first wave countries across advanced economies and China,GDP per capita growth could slow by 0.4 percent annually on average from 2023 to 2050,and up to 0.8percent in some countries,unless productivity growth increases by two to fo

15、ur times or people work one to five hours more per week.Retirement systems might need to channel as much as 50 percent of labor income to fund a 1.5-time increase in the gap between the aggregate consumption and income of seniors.Later wave countries,take note.In confronting the consequences of demo

16、graphic change,societies enter uncharted waters.Absent action,younger people will inherit lower economic growth and shoulder the cost of more retirees,while the traditional flow of wealth between generations erodes.Long-standing work practices and the social contract must change.More fundamentally,c

17、ountries will need to raise fertility rates to avert depopulationa societal shift without precedent in modern history.3Dependency and depopulation?Introduction Families across the globe are having fewer and fewer children.In much of the world,fertility rates have fallen below the replacement rate re

18、quired to maintain a stable population,and,despite increasing longevity,some countries have already started to see population decline.Others may follow in the not-so-distant future.1Falling fertility rates shift the demographic balance toward youth scarcity and more older people,who are dependent on

19、 a shrinking working-age population.Longer life spans accelerate the shift.This phenomenon has begun to play out across advanced economies and China,where in three-fifths of countries annual deaths already exceed births.2 Emerging economies have more runway,but they face the need to get richer befor

20、e the demographic transformation sets in.Our current economic systems and social contracts have developed over decades of growing populations,in particular working-age populations that drive economic growth and support and sustain people living longer lives.This calculus no longer holds.A combinatio

21、n of higher productivity,more work per person,effective migration,and higher fertility rates can ensure global prosperity for the future.That said,no one of those levers alone will be enough,and each presents challenges.Bending the trajectory of the demographic shift will require society to rethink

22、existing systems for work and retirement in ways that may compel a change in our social contractno easy feat.This report first explores the demographic shifts driven by falling fertility rates,until recently a trend primarily of interest to demographers and actuaries but now a topic of global conver

23、sation.3 Chapters 2 and 3 provide a comprehensive analysis of the potential economic consequences of falling fertility rates combined with longer life spans.The first wave of aging economies has started to feel the impacts on economic growth,labor markets,consumption,and public finances.Later waves

24、of aging will hit emerging economies,which are expected to face a similar fate just one to two generations later.We conclude with implications and recommendations for policy makers,businesses,and society.While the global population last declined significantly during the bubonic plague of the Middle

25、Ages,which is believed to have killed roughly half of Europeans,the demographic shift we are living through today is in many ways a result of vast improvements in global health,welfare,and prosperity.4 Humanity has demonstrated incredible resourcefulness throughout its history,and no doubt will find

26、 opportunities to thrive amid the challenges that a worldwide change in demographics poses.4Dependency and depopulation?Icy Macload/Getty ImagesAs well-being and prosperity increase around the world,two outcomesfewer children and longer livesare reshaping global populations.Over the past several dec

27、ades,families have shrunk in size virtually everywhere.In much of the world today,the total fertility rate,which we refer to as the fertility rate,is below the replacement rate of 2.1,which is the number of children needed to replace their parents.5(Demographic terms used in this report are defined

28、in the glossary.)As a result,the global age mix is shifting.While many people call this phenomenon“aging,”in fact the declining number of young peoplea youth deficitis driving the bulk of the demographic shift,a phenomenon we explore in this chapter.While declining fertility rates and changing popul

29、ation patterns are occurring everywhere,a first wave of regions,generally higher-income ones,has already begun to experience the effects of the demographic shift over the past several decades.Later waves of the same challenge will wash over many emerging economies in the next one to two generations.

30、Falling fertility is reshaping populationseverywhere Today in more than half of the worlds countries,home to two-thirds of humanity,the fertility rate has dropped below the replacement rate of 2.1 children per woman.Globally,the fertility rate averaged 2.3 children per woman in 2023,just over the re

31、placement rate.6 Over the past quarter century,the fertility rate has declined in 90 percent of the worlds countries(see sidebar“What is the fertility rate?”).The story of collapsing demographics starts in Luxembourg,the first country the United Nations recorded as having a fertility rate below repl

32、acement in 1950,when it first started collecting data(Exhibit 1).The countrys fertility rate rebounded in the 1950s,however,making Serbia and Croatia,both part of Yugoslavia at the time,the first countries where fertility permanently dropped below the replacement threshold,in 1963 and in 1968,respec

33、tively.Within a year,fertility rates in Denmark,Finland,and Luxembourg had followed suit.None of these countries has had a fertility rate equal to or above replacement since then.Twenty years later,most countries in Advanced Asia,Europe,and North America had crossed the replacement fertility thresho

34、ld,and China joined them in 1991.Subsequently,fertility rates in countries at varying levels of economic development around the world have fallen below replacementin Thailand in 1989,Mexico in 2015,and India in 2019.Sub-Saharan Africa is the one region of the world today where fertility rates remain

35、 high and are likely to stay above the replacement rate beyond the next quarter century.1.The age of youth scarcity6Dependency and depopulation?Sidebar:What is the fertility rate?The word“fertility”has a different meaning in the vernacular compared with its use among demographers.Colloquially,fertil

36、ity often refers to whether a person is able to conceive a child.But demographers focus on fertility rates,a measure of the average number of children women have throughout their life.Measuring the average number of children per woman in any given generation is straightforward once that generation h

37、as passed childbearing age.For example,the calculation today is easy for any cohort of women born before about 1970.However,this reflects only old information rather than providing an indication of how many children families are having today or are likely to have over their lives.To characterize the

38、 fertility rate in any given year,demographers measure the total fertility rate,which is a so-called period metric.The United Nations measures total fertility rate as the average number of children that would be born alive to a woman during her lifetime,assuming she were to pass through her childbea

39、ring years conforming to the age-specific fertility rates in that year.1 Age-specific fertility measures the number of children that women in each 1 Handbook on the collection of fertility and mortality data,United Nations Department of Economic and Social Affairs,2004.2 Alternatively,new research m

40、easures the effective fertility rate,which accounts for child mortality and brings the replacement rate to 2.The research finds that a portion of declining total fertility rates compensates for higher survival ratesalthough since the 1990s,this has only been the case in Sub-Saharan Africa,where chil

41、d mortality rates remained high.See Anup Malani and Ari Jacob,A new measure of surviving children that sheds light on long-term trends in fertility,National Bureau of Economic Research working paper number 33175,November 2024.age bracket have in the reference period and is typically collected for fi

42、ve-year brackets ranging from 15 to 50 years.The total fertility rate may differ from what cohort fertility rates ultimately end up being for many reasonsage-specific fertility rates change over time,for example.Fertility rates could be underestimated if younger women today end up having the same nu

43、mber of children as younger women 15 years ago,only later.In this case,the measured fertility rate will fall today,even if the eventual number of children born per woman is the same.Conversely,fertility rates may be overestimated for todays childbearing cohort if the number of children they have in

44、later adulthood continues to fall relative to older cohorts.Nonetheless,the total fertility rate provides a set of harmonized estimates that can be compared across societies,traced over time,and used to predict population size.Population size is expected to remain flat if the fertility rate is at re

45、placement level,defined by the United Nations as approximately 2.1 children per woman.This represents the average number of children a woman would need to give birth to in order to ensure that she is replaced by a daughter who survives to childbearing age.The true replacement rate also varies across

46、 countries and may be slightly higher or lower based on differences in child mortality and sex ratios at birth.2In the long run and setting migration aside,the population of a society with a fertility rate above replacement is expected to grow,while the population of one with a fertility rate below

47、replacement is expected to shrink.However,this rule of thumb doesnt necessarily hold in the short term,since population size also depends on the number of women of childbearing age and life expectancy.There are myriad causes for fertility changes around the world,ranging from societal shifts to deep

48、ly personal decisions.In this report,we do not explore the causes of declining fertility rates but rather examine their potential economic consequences.For a discussion of the causes of declining fertility rates,see,for example,Pablo Alvarez,“What does the global decline of the fertility rate look l

49、ike?”World Economic Forum,June 2022;Nicholas Eberstadt,“The age of depopulation:Surviving a world gone grey,”Foreign Affairs,November/December 2024;and Alice Evans,“Why is fertility collapsing,globally?”The Great Gender Divergence,November 1,2024.7Dependency and depopulation?Exhibit 1202319501965199

50、519802010Below replacement rate2.11.513 4 5 6 786.8111111111111111Sub-Saharan AfricaEmerging AsiaIndiaLatin America and CaribbeanNorth AmericaWestern EuropeAdvancedAsiax.xCentral and Eastern EuropeMiddle East and North Africa1.64.42.52.42.01.81.41.41.11.0Greater ChinaLuxembourg:2.0Sub-Saharan Africa

51、Emerging AsiaIndiaCentraland Eastern EuropeWestern EuropeAdvanced AsiaChinaSub-Saharan AfricaEmerging AsiaMiddle East and North AfricaLatin America and the CaribbeanMiddle East and North AfricaIndiaLatin America and the CaribbeanNorth AmericaNorth AmericaCentraland Eastern EuropeWestern EuropeAdvanc

52、ed AsiaGreaterChinaNote:The boundaries and names shown on this map do not imply ofcial endorsement or acceptance by McKinsey&Company.1The United Nations reports that the status of borders in this area has not been agreed upon by the parties.Source:World Population Prospects 2024,United Nations;McKin

53、sey Global Institute analysisFertility rates are declining everywhere.McKinsey&CompanyTotal fertility rate by country,live births per womanTotal fertility rate by region,live births per woman,19502023First waveLater waves 01950 2023 502350235023502350235023502350235023 2 4 6 8 0 2 4 6 8Below replace

54、ment rate2.11.513 4 5 6 788Dependency and depopulation?At the same time,life expectancy has increased almost everywhere.7 But greater longevity explains just 20 percent of the change in the age profiles of populations in developed countries since 1960;falling fertility rates explain the rest(Exhibit

55、 2).Consider the combined impact of these two forces in Germany.From 1960 to 2021,26 million fewer babies were born,a number equivalent to 31 percent of the German population at the end of that period,than if fertility rates had remained constant at the 1960 level.Seven million more seniors,or about

56、 8 percent of the countrys population,were alive at the end of the period due to increased life expectancy over the same period.Net-net,the countrys population in 2021 was 23 percent smaller than it would have been had both fertility and life expectancy rates remained constant,equivalent to roughly

57、19 million fewer people.Japan is the only developed country in our research where life expectancy had roughly the same impact as fertility.This unusual pattern was due to two factors.First,Japan already had a very low fertility rate in 19601.98 compared with,for example,2.7 in the United Kingdom at

58、that time.And life expectancy at 65 increased more in Japan than in other countriesby nine years compared with six years in the United Kingdom.In emerging economies,fertility rates fell even more dramatically from 1960 to 2023.For instance,a woman had an average of 6.1 babies in Brazil in 1960,where

59、as today,she has 1.6 children.Exhibit 2United StatesUnited KingdomSpainGermanyGreaterChinaAustraliaItalyFranceJapanSouthKorea870%100%852630308813810169391876314427151578382817978777250951Seniors defned as individuals aged 65 years and older in 2021.Source:World Population Prospects 2024,United Natio

60、ns;McKinsey Global Institute analysisFalling fertility rates explain about 80 percent of the changes in todays total population resulting from the demographic shift sweeping the world.McKinsey&CompanyImpact of changes in fertility rate and life expectancy on total population,19602021,%of 2021 popula

61、tionFEWER YOUNG PEOPLEAdditional people who would have been born had fertility rates not declined,%of 2021 populationShare of demographicshift explained by declining fertility,%MORE SENIORSAdditional seniors who would not be alive today had life expectancy not increased,%of 2021 population19Dependen

62、cy and depopulation?Global life expectancy has extended by seven years on average since 1997,reaching 73 years in 2023 and set to hit 77 years by 2050.Centenarians,or those 100 years and older,are the fastest-growing age group in percentage terms,according to the United Nations.Yet for all the atten

63、tion paid to rising longevity,declining fertility more powerfully determines global demographics.Populations morph from pyramids into obelisks Due to the demographic shifts weve described,what demographers call population pyramids are shaped less and less like pyramids today.To analyze how populatio

64、n structures are shifting over time,weve grouped the worlds countries into ten regions:Advanced Asia,Central and Eastern Europe,Emerging Asia,Greater China,India,Latin America and the Caribbean,Middle East and North Africa,North America,Sub-Saharan Africa,and Western Europe(for more on our classific

65、ation and data,see sidebar“Pyramid foundations:Demystifying the approach”).In most regions,these structures now resemble shallots,and in more economically advanced ones,they are taking on the shape of obelisks(Exhibit 3).Exhibit 3Later wave regionsFirst wave regionsPopulation breakdown by gender and

66、 age group,19602100Web MGI-DemographicsExhibit of Population pyramids will turn into obelisks by 2100.04101420243034404450546064707480849094100+Advanced AsiaCentral and Eastern EuropeWestern EuropeNorth AmericaAgeGreater China04101420243034404450546064707480849094100+EmergingAsiaLatin America and Ca

67、ribbeanSub-SaharanAfricaIndiaAgeMiddle East and North Africa1960202320502100 PopulationPopulationMaleFemaleMaleFemaleNote:Pyramids are drawn to scale within each region but not between regions.Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisMcKinsey&Company10D

68、ependency and depopulation?Societies are shaped in large part by their age structures,and their economic priorities shift as their population pyramids invert.The amount of physically intensive work done versus the number of physical therapists needed,or the number of families adding nurseries to the

69、ir homes rather than adding accessory dwelling units to house aging parents,depends on the mix of younger and older people in a population.Crucially,pension entitlements kick in between 60 and 67 years of age in many economies,particularly in advanced ones.At that point,seniors become recipients of“

70、support”provided by the working-age population,those aged 15 to 64,who generate most of the income and pay most of the taxes that support older people.Without significant changes,the worlds aging population means a growing number of older people who arent working will require the support of a shrink

71、ing number of younger people who are.Even if global fertility rates were to jump overnight to the replacement rate,it would take 20 years,give or take,for those additional babies to become adults and begin contributing to economic growth through work.Sidebar:Pyramid foundations:Demystifying the appr

72、oachIn this report,we explore the implications of changing demographics by analyzing projected population shifts based on United Nations forecasts.We examine the outcomes across different waves to understand the impact over various time horizons.We have consolidated the world into ten regions to pre

73、sent our findings,and each of the 237 countries and areas in the UN World Population Prospects 2024 is individually accounted for so that global populations sum to the total.Our first wave regions include Advanced Asia(Australia,Japan,New Zealand,Singapore,and South Korea),Western Europe,Greater Chi

74、na,Central and Eastern Europe,and North America.Later waves encompass Latin America and the Caribbean,India,the Middle East and North Africa,Emerging Asia(economies not included in Advanced Asia,India,or Greater China),and Sub-Saharan Africa.Throughout the report,we rely on 20 countries for examples

75、.For first wave regions,those countries are Australia,China,France,Germany,Italy,Japan,South Korea,Spain,the United Kingdom,and the United States.Together,these countries account for 28 1 See Nicole Maestas et al.,“The Effect of Population Aging on Economic Growth,the Labor Force,and Productivity,”A

76、merican Economic Journal:Macroeconomics,April 2023.percent of global population and generate 65 percent of the worlds nominal GDP.The example countries from later wave regions are Bangladesh,Brazil,Egypt,Ethiopia,India,Indonesia,Mexico,Nigeria,Thailand,and Trkiye.These countries account for 35 perce

77、nt of global population and 12 percent of nominal GDP.Most of our data sources are publicly available and standardized.The future population projections in this report are based on the medium scenario from the UN World Population Prospects 2024,which covers 237 geographic entities,mostly countries.T

78、he report includes population estimates from the 1950s to the present,as well as projections through 2100.Additionally,we have supplemented our analysis with income and consumption data from the National Transfer Accounts Project;labor market data from the International Labour Organizations ILOSTAT

79、and The Conference Board;and additional information from Eurostat,the OECD,and,where applicable,national statistical offices.We also use consumption data from World Data Lab and supplement our economic analyses with Oxford Economics data,which are not publicly available.Throughout the report,our ana

80、lysis takes a ceteris paribus approach,holding all variables constant except demographic change,to isolate its effect.Our approach has two key limitations to highlight.First,ceteris paribus outcomes throughout this report are not predictions;rather,they are thought experiments to explore to what deg

81、ree current economic systems“add up”in future demographic realities and to understand the ways in which they do not.These calculations are not dynamic,but we readily acknowledge that ceteris paribus never holds in practice.For example,we do not consider the impact of aging directly on productivity w

82、ithin countries,on personal savings rates,and so forth.1 Societies are highly adaptable,and they must and will adapt in ways that our calculations do not attempt to anticipate.In fact,the size of the required adjustment is what we attempt to approximate with our methodology.Second,our sources of dat

83、a,while robust,have limitations.In some instances,different sources do not agree,and some gaps in the data require interpolation.In some cases,data sources are themselves estimating quantities with a high degree of uncertainty.Even the UN population scenarios,while the closest assessments available

84、to a global consensus view,have their critics(see sidebar“Predicting the future is hard,and demographers dont agree”).11Dependency and depopulation?Sidebar:Predicting the future is hard,and demographers dont agreeThe complex interplay between fertility,life expectancy,migration,and existing populati

85、on structures throughout the world not surprisingly creates uncertainty about how populations will evolve in the future.Many organizations forecast trajectories,1“Definition of projection scenarios,”in World Population Prospects 2024,United Nations,July 2024.2 See“How does low fertility affect econo

86、mic growth,worldwide?Jess Fernndez-Villaverde,”Rocking our priors podcast,Alice Evans,October 2,2024.which often vary widely.The projections in this report are based on the UN medium scenario from 2024.1 The United Nations is keenly aware of the inherent uncertainty in its calculations:its low and h

87、igh scenarios for global population in 2100 differ by a factor of two,ranging from seven billion humans to 14billion humans(Exhibit A).Other well-established research centers such as the Institute for Health Metrics and Evaluation and the International Institute for Applied Systems Analysis use alte

88、rnative methodologies to estimate population projections,with marginally different results.While the UN medium scenario is the closest thing to a global consensus view,it is not without its critics.For instance,some contend that it is too optimistic in its forecast that fertility rates will stop fal

89、ling and slightly rebound in many countries(Exhibit B).2 Exhibit AGlobal population,million,by diferent UN scenarios and other data sourcesUN high scenarioIIASA1 medium scenario(SSP2)IHME2 reference scenarioUN medium scenario10,9359,6648,94214,39510,1809,8859,7866,987UN lowscenarioHistorical1Interna

90、tional Institute for Applied Systems Analysis.IIASA data are available for 200 countries,by age,sex,and education levels according to 7 scenarios from the Shared Socioeconomic Pathways(SSP)Scenario Database.These include scenarios numbered SSP1 through SSP5,and two additional scenarios which combine

91、 SSP2 with zero migration and with double migration.2Institute for Health Metrics and Evaluation.IHME data are available for countries and regions,by age and sex according to 5 scenarios.Source:World Population Prospects 2024,United Nations;IHME;IIASA;McKinsey Global Institute analysisThe UN medium

92、scenario refects the rough global consensus view of population trends,but projections vary across sources and scenarios.McKinsey&Company5,00010,00012,5007,500 15,000 1997 20232050210012Dependency and depopulation?Another important variable that UN projections cannot account for is changes to migrati

93、on patterns.While fertility and longevity move slowly,migration can change suddenly due to changing economic conditions,conflict,or climate,among other factors,causing large swings in the evolution and location of populations that are extremely hard to predict.All in all,we are confident that the di

94、rection of our insights is correct,useful,and relevant to policy makers,businesses,and individuals.At the same time,it is important to note that specific numbers quoted in this research may not reflect exactly how the future will unfold,and so the report should be read as a scenario to prepare for r

95、ather than a prediction.Sidebar:Predicting the future is hard,and demographers dont agree(continued)Exhibit B2.62.42.03.02.82.2World2000 05101520 20252.01.91.61.82.22.11.7United States2000 05101520 20251.41.20.82.01.81.61.0China2000 05101520 20251.51.41.11.21.71.61.3Japan2000 05101520 20251.41.20.61

96、.01.81.60.8South Korea2000 05101520 20251.41.21.01.81.6Spain2000 05101520 2025Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisThe UN medium scenario predicts the total fertility rate will plateau or even rebound in every release,but so far it keeps declining.M

97、cKinsey&CompanyFertility rate,real and UN medium scenario releasesUN medium scenario releases2006Real20082010201220152017Working-age populations will peak and fall in three wavesWorking-age people account for the bulk of economic output,so their numbers relative to those of older and younger people

98、determine a host of economic outcomes.All regions will see the share of working-age people in their populations decline,although at different paces and points in time.First wave regions are those already undergoing this change.Later wave regions,where the shift is just beginning to take hold or hasn

99、t yet arrived,will experience a peak and subsequent decline in the share of working-age population in the futurein some cases,the near future.13Dependency and depopulation?Among first wave regions are predominantly developed economiesAdvanced Asia,Central and Eastern Europe,North America,and Western

100、 Europeand Greater China,which has lower GDP per capita than other first wave regions but shares their demographic characteristics.8 These regions have an average total fertility rate of 1.2 children per woman today,and 67 percent of their combined population is working age,down from a high of 70 pe

101、rcent in 2010(Exhibit 4).In aggregate,this cohort is rapidly shrinking in these regions,where the share of the working-age population is projected to drop to about 59 percent by 2050.There are two later wave groups of regions.A second wave has just reached the shores of Emerging Asia,India,Latin Ame

102、rica and the Caribbean,and the Middle East and North Africa.Their total fertility rate is 2.2,and 67 percent of their population is working age today.This wave is still gathering momentum,however,and will peak in the 2030s in aggregate.In Sub-Saharan Africa,the average fertility rate is 4.4 today,an

103、d just 56 percent of the population is working age.This share will continue to grow,peaking at 66 percent well into the second half of the century,when the third wave of the demographic shift hits its shores.The three waves mask variations,and regions,countries,counties,and cities within them are at

104、 different stages of the demographic shift.For example,in the first wave,Germanys working-age population share peaked in 1986,while in the United States and China,it peaked in 2007 and 2010,respectively.China is projected to experience a sharper decline in its working-age population share in the fut

105、ure,given its lower fertility rate and lower migration.The United States and India measure fertility rates at a subnational level,providing examples of how birth rates vary even more within countries.(See sidebar“Fertility rates vary significantly across regions and within countries.”)Exhibit 4Popul

106、ation aged 1564 years,%of total population Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisWorking-age populations peak in three waves.McKinsey&Company80757065605055451960First waveLater waves(excluding Sub-Saharan Africa)Sub-Saharan AfricaToday197019801990200

107、0201020202030204020502060207020802090210014Dependency and depopulation?Sidebar:Fertility rates vary significantly across regions and within countriesSome large nations,such as the United States and India,monitor the total fertility rate at a subnational level,capturing this data by state(exhibit).In

108、 the United States,the overall fertility rate is 1.66,according to the US Centers for Disease Control and Prevention.Rates vary by state:South Dakota has a fertility rate of 2.01,just slightly below the replacement level and comparable to that of Indonesia.Texas,the second-most-populous state,has a

109、fertility rate of 1.84.By contrast,Vermont and Washington,DC,have fertility rates of 1.35 and 1.24,respectively,figures that are more comparable to those of Finland or Switzerland.Fertility rates vary even more widely in India.While the national fertility rate stands at 1.98,below the replacement le

110、vel,about a third of the population resides in states where fertility rates exceed replacement levels.Sikkim,the state with the lowest fertility,is at 1.05,according to the fifth National Family Health Survey.Conversely,Bihar has a fertility rate of 2.98,between Kenya and Botswana.Even within Indian

111、 states,fertility rates vary among populations,reflecting urban and rural differences.In Bihar,the urban fertility rate is 2.35,while the rural fertility rate is significantly higher,at 3.11.Thus,while fertility is a national issue for many countries,a subnational approach to addressing falling fert

112、ility rates may yield richer results.ExhibitUnited States,20221.662.01South Dakota1.24District of Columbia2.00India,2019212.98Bihar1.05SikkimTotalTotalThe boundaries and names shown on this map do not imply ofcial endorsement or acceptance by McKinsey&Company.Source:Centers for Disease Control and P

113、revention;National Family Health Survey(India);McKinsey Global Institute analysisFertility rates vary signifcantly even within individual countries.Total fertility rate by state,live births per womanMcKinsey&CompanyBelow replacement rate2.01.81.61.42.1 2.4 2.6 2.8315Dependency and depopulation?Never

114、theless,the broad momentum and peak of this population cohort in countries within each wave are similar,reflecting the horizons over which they will confront youth scarcity.For first wave regions,the declining share of working-age population is a relatively new development,and many companies,governm

115、ents,and communities havent yet fully come to grips with the implications.Later wave regions,excluding Sub-Saharan Africa,still have time to prepare,but not much.Support ratios will continue to fallAs the world becomes youth scarce,the number of workers per senior will fall.To illustrate this,we foc

116、us on the support ratiothe number of people aged 15 to 64 years,or those of working age,relative to the number 65 years and older.Globally,the support ratio was 9.4 in 1997.Put differently,there were more than nine working-age people to support one older person(Exhibit 5).Exhibit 5Sub-Saharan Africa

117、Emerging AsiaLatin America and CaribbeanWestern EuropeGreater ChinaIndiaAdvancedAsiaSupport ratio,number of people 1564 years old for each person 65 years and older,19972050Support ratios will decline rapidly across regions through 2050.018128421416106Central and Eastern EuropeNorth AmericaTotal frs

118、t wave regionsTotal later wave regionsTOTAL WORLDMiddle East and North AfricaSource:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisMcKinsey&CompanyFirst wave regionsLater wave regions199720232050By 2050,the support ratio is expected to fall to 3.9that is,fewer than

119、four people will need to support each senior.16Dependency and depopulation?Today,the global support ratio is 6.5,and by 2050,it is expected to fall to 3.9that is,fewer than four people to support each senior.This trend is starker in first wave economies,where the support ratio is already 3.9 today,d

120、own from 6.8 in 1997.The ratio is expected to fall to two working-age individuals for every person over 65 years by 2050.Among regions in the first wave,Advanced Asia,Greater China,and Western Europe will have the lowest support ratios by 2050;the ratio will fall fastest in Greater China.In later wa

121、ve regions,the support ratio will decline from 10.3 today to 5.7 by 2050.India,where it is 9.8 today,offers a particularly stark example.In the country currently home to the worlds largest population,the trajectory of birth rates and life expectancy indicates its support ratio will be roughly half w

122、hat it is today by 2050 and drop to 1.9about the same as Japan todayby 2100.First wave countries populations have already peaked The world reached its maximum number of annual births in 2012,when 146 million babies were born,and the global number of births will continue to slowly decline.According t

123、o the United Nations,the total number of people on Earth will peak in 2084,at just above ten billion,and start declining in the latter years of this century.Total population in first wave regions,however,peaked in 2020.On the current trajectory,the population of these regions will fall from 2.8 bill

124、ion today to 2.6 billion by 2050 and to 1.9 billion by 2100(Exhibit 6).Only 22 of the 55 countries in these regions will have more people in 2050 than today,and populations in most of those countries will decline thereafter.9 Already,more people die each year than are born in 37 countries in first w

125、ave regions.Today,60 percent of the worlds population aged 65 and older resides in these regions.By contrast,only 22 percent of those younger than 15 years live there.Populations across later wave regions are still increasing.The second waves total population will reach its maximum by 2071,going fro

126、m four billion today to five billion at its peak.Sub-Saharan Africas population will still be growing by the turn of the century and is projected to reach 3.5 billion by then,up from 1.3 billion today.(See sidebar“Predicting the future is hard,and demographers dont agree.”)17Dependency and depopulat

127、ion?The population balance will shift decisively to later wave countriesThese dynamics mean that the planets population is shifting toward later wave regions.By 2050,a quarter of the global population will live in first wave regions,compared with 35 percent of the worlds people today.According to UN

128、 projections,these regions could be home to less than 20 percent of the global population by 2100(Exhibit 7).Even though Sub-Saharan Africas fertility rate is falling fast,almost 300 of the worlds next thousand babies will be born there.Nigeria alone will become home to 57 of the next thousandor fiv

129、e more than the 52 born across Central,Eastern,and Western Europe combined.Similarly,172 of the next thousand babies the stork delivers will be in India,where the birth rate overall has dropped below replacement but where the current population of women of childbearing age is still high.By 2100,Sub-

130、Saharan Africa will account for all of the net global population increase,doubling its current share to 34 percent.10 By contrast,Greater Chinas share of the global population,today the second largest among the ten regions,will shrink by two-thirds,from 18 percent in 2023 to 6 percent by 2100.This w

131、ould make Greater Chinas population only 170 million larger than North Americas,according to UN estimates,compared with a difference of roughly one billion people today.Exhibit 6Later wave regions excluding Sub-Saharan Africa1997 202320502100012 345Sub-Saharan Africa1997 202320502100012 34Total worl

132、d 65 years and older15-64 yearsUnder 15 years01997 20232050210012 3456 789101110.29Peak20842023Peak207120238.09+27%3.53Peak210020231.29+174%+25%Peak202020232.812.82First wave regionsAGE GROUP01997 20232050210012 3Population,19972100,billion Source:World Population Prospects 2024,United Nations;McKin

133、sey Global Institute analysisWorld population is projected to peak at more than ten billion in 2084,but populations in frst wave regions have already peaked.McKinsey&Company4.085.1118Dependency and depopulation?Labor markets are expected to shift accordingly to young and growing countries in later w

134、ave regions unless migration patterns change dramatically,a trend we explore in greater detail in subsequent chapters.By 2050,holding current hours worked per capita constant within each age group,later wave regions would account for more than two-thirds of all hours worked globally.At that time,Sub

135、-Saharan Africa alone could account for 18 percent of global hours worked,doubling its share of work hours today.The share of the worlds work done by Chinese workers,on the other hand,could drop to 18 percent by 2050 from 26 percent today,and every other first wave regions share is set to shrink.11

136、This could create an opportunity for many later wave countries to progress economically.Opportunities span the entire tradeable economyservices as well as manufacturing.At least for the next quarter century,countries in later wave regions will account for more than half of global consumption,too,due

137、 to fast-growing young populations and growing incomes.For example,World Data Lab projects that India and Emerging Asia will account for 30 percent of global consumption at purchasing-power parity(PPP),up from 12 percent in 1997.By comparison,Advanced Asia,North America,and Western Europe could acco

138、unt for just 30 percent of the worlds consumption then,down from 60 percent in 1997(Exhibit 8).12Exhibit 7Share of total population by region,%Total population,millionGreater China42%35%26%18%Advanced AsiaNorth AmericaWestern EuropeCentral and Eastern EuropeLatin America and CaribbeanMiddle East and

139、 North AfricaEmerging AsiaIndiaSub-Saharan AfricaWeb Exhibit of Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisBy 2100,less than 20 percent of the worlds population will live in frst wave regions.McKinsey&Company19972023205021005,9258,0929,66410,1802118136332

140、554575464388866789161617181718171511162334142Share acrossfrst wave regionsFirst wave regionsLater wave regions(excluding sub-Saharan Africa)Sub-Saharan Africa19Dependency and depopulation?This shift has important implications for many businesses in first wave and later wave regions alike as they dev

141、elop and scale products to serve growing markets.As income and consumption grow in later wave regions,local and multinational companies will need to determine how to best meet changing local tastes and adapt their products and services accordingly.Affordability may increase in importance as consumer

142、s in later wave regions present a bigger market opportunity for businesses.Navigating as a business may become more complex,as many later wave countries have more challenging legal and governance environments and are more conflict prone.13Prospects of depopulation put the focus on fertility rates Ov

143、er the longer term,countries in first wave regions may face the challenge of depopulation(Exhibit9).14 Populations in 26 countries in these regions are on track to decline by a third or more by 2100,while in countries including China,Poland,and South Korea,they are expected shrink by half or more.Pr

144、ojections suggest that some countries with fertility rates below replacement,including France,the United Kingdom,and the United States,will have continued population growth through 2100 based on positive net migration.Smaller populations could affect many sectors of the economy as the number of chil

145、dren born dwindles.If a country had a constant fertility rate of 0.7,equivalent to South Koreas rate today,it would have just 13 grandchildren per 100 grandparents two generations from now.Just imagine school systems needing to adjust to a precipitous drop in the number of students,resulting in clos

146、ing schools and extended travel distances.Declining populations would also challenge debt sustainability and the social contract,not to mention the global geopolitical balance.While some contend that smaller populations could reduce carbon emissions,MGI research has found that Exhibit 8Global consum

147、ption composition,$trillion 2021 PPPRegional contribution to global consumption,%,19972050Note:Consumption estimates from World Data Lab include both demographic efects and projected increases in per capita consumption(income efects).Only direct consumption is included;in-kind consumption from gover

148、nments is excluded.Figures may not sum,because of rounding.Source:World Data Lab;McKinsey Global Institute analysisConsumption pools are shifting from North America and Western Europe to Emerging Asia and India.McKinsey&CompanyEmerging AsiaLatin America and CaribbeanIndiaMiddle East and North Africa

149、Sub-Saharan AfricaNorthAmerica36+48+9619972023205019972050231997205023179Western EuropeGreaterChinaCentral and Eastern EuropeAdvancedAsiaFirst waveLater waveSub-Saharan Africa89811053251028318426675117524149413132217710156781087345491620Dependency and depopulation?sustained economic growth is essent

150、ial to pay for the net-zero transitionand this may come into question if populations shrink.How different would first wave regions look in 2050 and 2100 if fertility rebounded?We developed what we call the“replacement floor hypothetical,”which assumes that countries with fertility rates below the re

151、placement rate today achieve a replacement-level fertility rate of 2.1 births per woman starting in 2024 and for every subsequent year.While this outcome is highly unlikely,it is an informative way to think about an alternative future.Changing fertility rates dont have an immediate effect,although d

152、ue to compounding,having more children today under any scenario would have a big impact over the longer term.Under the UN medium scenario,the global population would reach 9.6 billion by 2050,while it would climb to 10.2billion under our replacement floor hypothetical over the same period.By 2100,po

153、pulations under the two scenarios would grow to 10.2 billion and 12.6 billion,respectively.In Advanced Asia,achieving our hypothetical would lead to a population of 247 million in 2100,more than 1.6 times the population expected under the UN medium scenario(Exhibit 10).Exhibit 9Population decline,21

154、00 versus 2023,%Population,2100,million Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisBased on current projections of fertility and longevity,many countries are headed toward population collapse by 2100.First wavecountriesFirst waveregionsLater wavecountries

155、McKinsey&Company Greater China Advanced Asia Central and Eastern Europe Western Europe North AmericaChinaItalyJapanGermanyFranceUnited KingdomUnited States6461522303784756333577716874421651631,50529647756%16%58%50%41%38%31%Decline,2100 vs 2023,%100%0%80%60%40%20%56%+24%+23%+3%+8%55%41%38%16%Populati

156、on expected to increase by 210031%TrkiyeBrazilIndiaIndonesia38%+5%+5%25%38%23%Nigeria+109%12%29%21Dependency and depopulation?Support ratios would also remain significantly higher,although still lower compared with today.By 2100,support ratios in first wave regions would level out at about 2.2 peopl

157、e of working age per senior,compared with a ratio of 1.4 under the UN medium scenario.In Western Europe,for example,the support ratio would increase from a projected 1.7 under the UN scenario to 2.2 under our hypothetical.Increasing fertility rates,which traditionally fall as national income and fem

158、ale labor force participation rates rise,can be challenging.There are no clear examples of countries successfully boosting their birth rates significantly,although many are trying.For instance,the South Korean government has underwritten postpartum care centers in an effort to make childbirth as tro

159、uble-free as possible.15 Hungary offers one-time cash incentives to new parents and spends almost 6percent of its GDP on fertility and family programs overall.16 Several Western European countries have extended family leave policies;Norway,for instance,covers the incomes of new parents up to a maxim

160、um of 49 weeks set by its welfare system,and many Norwegian employers choose to cover any gap in income to maintain full salaries.17 Despite these efforts,none of these countries has managed to push fertility rates back to the replacement level,although each initiative offers insights about what doe

161、s and does not work to influence fertility rates over the longer term.18 Research tracking these efforts suggests that,on their own,policies that have been implemented and evaluated in high-income countries to date are unlikely to lead to substantial or sustained increases to the birth rate.19Regard

162、less,any children born over the next few decades will not enter the workforce right away.Thus,even as countries consider steps to address population decline in the long term,they need to adapt to demographic shifts over the next two decadesthe die has already been cast.Exhibit 10North AmericaCentral

163、 andEastern EuropeGreater China2050210020502100205021002050210020502100WesternEuropeAdvanced Asia1,4971,2862.11.92.01.02.82.62.52.02.01.92.21.72.52.32.61.91.71.52.11.56464274604754254653785492973282303571952191522476341,384McKinsey&CompanyWeb Exhibit of Total population,millionUN medium scenarioRepl

164、acement foor hypothetical1Support ratio,number of people 1564 years for each person 65 years and older1“Replacement foor hypothetical”is a case we developed that assumes countries with fertility rates below the replacement rate of 2.1 children for every woman in 2024 immediately achieve that fertili

165、ty rate.Source:World Population Prospects 2024,United Nations;McKinsey Global Institute analysisRaising fertility rates today to match the replacement rate could increase populations and support ratios by 2050 and 2100.22Dependency and depopulation?kali9/Getty ImagesMany first wave economies face a

166、virtually unprecedented depopulation challenge toward the end of the century,according to UN projections.More immediately,they face another challenge:increasing dependency that could depress economic growth over the next quarter century.As population pyramids become bottom-light and top-heavy,the we

167、ll-being of a growing legion of older people and of society at large will depend on a stagnant or shrinking number of people who work,which will increase pressure on public finances.Youth scarcity could also modify consumption and savings patterns.In this chapter,we begin by analyzing the consequenc

168、es of the demographic shift on labor markets and on GDP per capita growth by 2050 in first wave regions(chapter 2.1).Our findings are not projections:we have taken an all-else-being-equal approach,keeping all variables constant except the demographic structures,or age mix,of economies.A combination

169、of three levershigher labor intensity,more robust productivity growth,and shifts toward a younger age mix via higher fertility and effective migrationcan offset the headwinds of the youth deficit(chapter 2.2).In many countries in first wave regions,however,no one of those levers alone can sustain pa

170、st GDP per capita growth rates.We then examine what could happen to savings trends and public finances(chapter 2.3).Finally,we examine changes in labor and consumption patterns that businesses may face and consider how they can respond(chapter 2.4).2.A lower-growth,higher-dependency future in first

171、wave regions24Dependency and depopulation?2.1.Demographic shifts,slower economic growthPeople typically work less as they get older,so the shifting demographic trends weve mapped could slow economic growth.GDP per capita depends on the number of hours worked per person and how productive each hour o

172、f work is,or productivity.Hours worked,in turn,depend on how much individuals of each age work,or labor intensity,and the number of people in each age group,or the age mix.Thus,GDP per capita growth depends on productivity growth,shifts in the age mix,and growth in labor intensity among people in ea

173、ch age cohort.Under current projections,a changing age mixmore older people and fewer working-age peoplewill result in slower growth in hours worked and thus reduce GDP per capita growth if left unaddressed.In this section,we analyze the consequences of this shifting age mix on the growth in hours w

174、orked per person.Throughout,we assume that labor intensity in each age group remains constant at 2023 levels(see sidebar“Our approach to sizing the impact of demographic changes and how to counteract them”).To maintain GDP per capita growth,countries will need to influence their age mix,increase lab

175、or intensity,or boost productivity growthor,more likely,rely on a combination of all three.In the following section,we analyze how much these three levers would need to be pulled in order to maintain GDP per capita growth at the same rate recorded over the past 25 years.2.125Dependency and depopulat

176、ion?Sidebar:Our approach to sizing the impact of demographic changes and how to counteract themGDP per capita growth depends on growth in productivitythe value of output produced per hour of workand growth in hours worked per person.The latter,in turn,depends on labor intensity growth in each age gr

177、oup and changes in the age mix.Labor intensity is driven by the share of the population that seeks work(the labor force participation rate),the share of those who find employment,and the number of hours each worker works.The age mix depends on fertility rates,longevity,and migration flows.Productivi

178、ty growth is driven by physical and human capital investments as well as the rate of innovation.Take Western Europe as an example(exhibit).As more women and older people joined the labor force there,labor intensity grew by 0.4 percent on average annually from 1997 to 2023.However,as the regions age

179、pyramid skewed older,primarily due to falling fertility rates,the changing age mix reduced economic growth by 0.3 percent per year.Productivity grew by 0.8 percent per year.Combined,the three components brought overall GDP per capita growth to 1percent a year.As the number of older people whose labo

180、r intensity is low increases and middle-aged cohorts with high labor intensity shrink,the age mix will increasingly depress growth,absent changes.Thus,labor intensity or productivity will have to grow faster to maintain past GDP per capita growth,or the age mix will need to shift in ways that differ

181、 from current UN projections via changes in fertility rates or migration.We aim to answer two questions in the first two sections of chapter 2:1.How much will the change in the age mix projected under the UN medium scenario decrease growth in hours worked per capita and thus GDP per capita growth(ch

182、apter 2.1)?2.What changes in productivity,labor intensity,or the age mix would be necessary to maintain past GDP per capita growth(chapter 2.2)?In chapter 2.1,we size the impact of the age mix on the growth of hours per capita.We hold the other component that drives hours,labor intensity per age gro

183、up,constant at 2023 levels.In chapter 2.2,we consider target GDP per capita growth,which is the average annual growth in a country or region over the past 25 years,and size the changes needed in productivity growth and labor intensity to maintain that growth,given the drag of the baseline age mix ca

184、lculated in chapter 2.1.In only two countries does the target we use differ from their average 19972023 GDP per capita growth:China and South Korea.In 1997,these countries were emerging economies,so targeting their high past growth would be unrealistic.Instead,we calculated new targets for them base

185、d on the growth trajectory of economies that developed earlier.The GDP per capita ExhibitSource:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisWhat drives GDP per capita growth?McKinsey&CompanyFactors driving GDP per capita growth,199720

186、23GDP per capita,$Labor force participation Employment rate Hours per worker Fertility Longevity Migration Human capital Physical capital Technology,innovation Job mixExample:Western Europe,GDP per capita,CAGR 19972023,%FactorsDrivers02000 20242050 10,000 20,000 30,00040,000 xLabor intensityHow much

187、 each individual worksAge mixRelative number of people per age groupProductivityHow productive each hour of work is0.30.40.81.026Dependency and depopulation?Labor intensity declines at older agesbefore official retirement Labor intensity,measured as weekly hours worked per person in an age group,sta

188、rts declining everywhere before official retirement ages.Across first wave regions,weekly hours worked per capita peak at about 50 years of age and decline thereafter(Exhibit 11).The primary reason is falling labor force participation ratesfewer older people continue to workbut on average,older work

189、ers who are employed also work fewer hours.In Germany,for instance,people aged 25 to 54 years work 25 hours per capita a week,on average,while those aged 55 to 64 years work 21 hours a week.20 Among Germans who are 65 years and older,this plummets to an average of two hours per capita per weekthat i

190、s,two hours per person,including those who arent in the labor force.The rapid decline of hours per capita reflects two trends:lower participation in the labor force and less intensity among those who do workbut the net effect is what matters for an economy.Japan,a society with more older people,has

191、more older workers,but labor intensity there still declines with age:Japanese between 25 and 64 years work 30 hours on average,and those 65 years and older work seven hours(see sidebar“Learning from Japan”).growth target for China is 4.9 percent per year,while that of South Korea is 2.0 percent.Many

192、 combinations of labor intensity and productivity growth could result in the same GDP per capita growth,so we show two of these combinations:the required increase in labor intensity if past productivity growth remains constant,and the required annual productivity growth if past hours per capita grow

193、th remains constant.Returning to the example of Western Europe,average GDP per capita growth in the past quarter century was 1 percent per year,which we use as the target for future growth.The regions age mix will drag growth down by 0.4 percent a year.Holding past productivity growth constant at 0.

194、8 percent,labor intensity would need to grow by 0.6percent annually to meet the GDP per capita growth target.By contrast,holding hours per capita growth constant at 0.2 percent,future productivity needs to grow by 1.2 percent annually to meet the target.Additional migration changes the age mix and t

195、hus affects average weekly hours worked per capita,so we separately calculate how many additional migrants would be needed to reach the required increase in hours,assuming past productivity growth stays constant.Sidebar:Our approach to sizing the impact of demographic changes and how to counteract t

196、hem(continued)27Dependency and depopulation?Exhibit 11Average frst wave regions150 years and olderAdvanced AsiaNorth AmericaCentral and Eastern EuropeGreater ChinaWestern Europe403020100501519252935394549555965+1519252935394549555965+1519252935394549555965+403020100501519252935394549555965+151925293

197、5394549555965+1519252935394549555965+Age groupNote:Hours per capita takes into account the entire population in each age group,not just workers.1Simple average of frst wave regions.Source:ILOSTAT;World Population Prospects 2024,United Nations;McKinsey Global Institute analysisLabor intensity starts

198、declining when workers are in their 50s.McKinsey&CompanyWeekly hours per capita over a life cycle,202328Dependency and depopulation?Sidebar:Learning from JapanCountries within waves are not all aging at the same pace.Japans fertility rate fell earlier than the rates of its first wave peers,and its a

199、verage life expectancy at birth is the highest in the world at 84 years.In 2005,20percent of its population was 65 years and older,and today that age cohort makes up 30 percent of the Japanese population.Projections suggest that the United States will be as old as Japan is today after 2100.France co

200、uld reach this milestone in 2079,China in 2047,and Spain in 2037.This makes Japan an interesting case study and raises 1 For a discussion of the impact of demographics on Japans economic growth,see Jess Fernndez-Villaverde,Gustavo Ventura,and Wen Yao,The wealth of working nations,National Bureau of

201、Economic Research working paper number 31914,November 2023.the question:What can other first wave countries learn from Japan?The first lesson is that as support ratios decline,labor intensity must risefewer working-age people means that,all else being equal,others may need to work more to sustain ec

202、onomic growth.Japans labor intensity is higher than that of virtually all other developed countries,and even more so at older ages(exhibit).The labor force participation rate among Japanese aged 50 to 65 years is 84 percent,up from 73 percent in 1997.Among people 65 years and older,it is 26 percent,

203、compared with 19 percent in the United States and 4 percent in France.The second lesson Japan offers is that higher labor intensity alone may not be enough.Japans annual labor productivity has grown 1.1 percent per year,on average,since 1997,faster than the 0.8 percent increase in Western Europe and

204、 somewhat less than the 1.5 percent growth in the United States.Yet Japans GDP per capita has grown just 0.6 percent,compared with 1 percent in Western Europe and 1.4 percent in the United States.1 The demographic drag is inexorable and severe,and when it hits,boosting productivity growth becomes ev

205、en more relevant.ExhibitWeekly hours per capita over a life cycle,2023Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisJapanese people work more hours,on average,than people in most other frst wave economies.McKinsey&Company15192024

206、2529303435394044454950545559606565+JapanOthers(Australia,Germany,Italy,Spain,United Kingdom)Age group,years20 1510 5025303540ChinaUnited StatesSouth KoreaFrance29Dependency and depopulation?Fewer hours worked will depress economic growthBecause labor intensity declines with age,a future with fewer y

207、oung and middle-aged people and more older people will mean lower growth in hours per capita per week,or labor intensity,all else being equal.From 1997 to 2023,the cohort aged 65 years and older was the fastest-growing age group in every first wave region.But the number of working-age people also in

208、creased in many first wave countries.The number of children younger than 15 years fell,a reflection of declining fertility rates.This combination helped partially balance the age mix for a while,diminishing the impact of increasingly aged populations.However,over the coming 25 years,the number of ol

209、der people living in first wave regions will continue to grow while every other age cohort shrinks.This shift in the age mix could slow the growth in hours worked per capita across first wave regions by 2.2 hours per capita per week on average,thus slowing GDP per capita growth(Exhibit 12).Exhibit 1

210、2Age groupWeb Exhibit of Note:Figures may not sum to 100%,because of rounding.Source:ILOSTAT;World Population Prospects 2024,United Nations;McKinsey Global Institute analysisDemographic headwinds will be a drag on hours per capita.McKinsey&CompanyContribution to hours per capita decline by age group

211、,frst wave economiesPopulation,millionThe change in.Change in total hours per capitaHours per capita per week,2023 0 0 10 2040 3080040001445231548733861050578469948177315293044456465Total change2.2Change in contribution to total hours per capita per week,202350 x=00.80.90.70.230Dependency and depopu

212、lation?From 1997 to 2023,the age mix effect was neutral or slightly negative across first wave regions,dragging down hours of work per capita by a negligible amount.Over the next quarter century,the age mix will slow the growth in hours per capita per weekand thus GDP per capita growthby 0.4percent

213、annually in every first wave region other than North America(Exhibit 13).This is equivalent to a cumulative reduction in hours per capita per week of 2.2 hours across first wave economies from 2023 to 2050,ranging from 0.8 to 2.6 in different regions.A 0.4 percent drag on GDP per capita growth per y

214、ear may seem trifling,but it isnt.The shift in age mix could slow GDP per capita growth over the next quarter century,for example,by an average of$10,000 in Western Europe and$6,000 in Greater China.Exhibit 13199720230.00000.20.20.10.20.20.40.40.30.70.00.40.40.40.30.40.40.10.80.420235019972023202350

215、SpainAdvanced AsiaBY COUNTRY1997202320235019972023202350AustraliaNorth America1997202320235019972023202350United KingdomWestern Europe199720230.20.10.10.10.40.20.30.7202350France19972023202350Germany19972023202350United States19972023202350Italy199720232023501997202320235019972023202350JapanGreaterC

216、hina19972023202350South KoreaCentral and Eastern EuropeContribution of age mix shift to hours per capita growth,19972023 vs 202350,%Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisDemographic shifts will slow GDP per capita growth

217、across frst wave economies.McKinsey&Company31Dependency and depopulation?Naturally,regional averages hide country variation.For instance,the effect of the demographic shift is already apparent in Australia,where a changing age mix has decreased weekly hours and GDP per capita growth by an average 0.

218、2 percent per year over the past quarter century,a decrease forecast to continue at the same pace to 2050.Spain offers a particularly extreme example of the age mix effect.While it was neutral from 1997 to 2023,age mix could deduct 2.8 hours per capita per week,or 0.8 percent of annual GDP per capit

219、a growth to 2050.More women and older people workingand productivityhave supported past growthBefore turning to changes needed in future labor intensity and productivity growth if GDP per capita is to be maintained,it is worth asking how they have evolved in the past.Aggregate labor intensity increa

220、sed in Western Europe from 1997 to 2023 and declined in Greater China and Advanced Asia(Exhibit 14).It remained flat on average in North America and Central and Eastern Europe.But the labor intensity of two groupswomen and older peoplegrew consistently in almost everyregion.Those 50 years and older

221、increased their participation in labor markets everywhere.This effect was most pronounced in Europe but apparent in every region,even in Advanced Asia,where labor intensity overall declined.Second,women aged 25 and olderparticularly older womenworked more.This was true in every region except Greater

222、 China,where womens labor intensity did not change.In Advanced Asia overall,labor intensity grew,though it fell slightly among women aged 35 to 50 years.Despite increased labor intensity among those older than 50 years and women in most first wave regions over the past quarter century,most regions h

223、ad stagnant or falling aggregate labor intensity,largely because fewer people younger than 29 years worked.Western Europe was the only first wave region where overall labor intensity not only increased but more than completely offset the negative impact of the changing age mix.32Dependency and depop

224、ulation?Exhibit 14Note:Hours per capita takes into account the entire population in each age group,not just workers.Source:ILOSTAT;World Population Prospects 2024,United Nations;McKinsey Global Institute analysisMcKinsey&CompanyWeekly hours per capita over a life cycle,by sex,1997 and 2023Total frst

225、 wave regionsAdvanced AsiaNorth AmericaCentral and Eastern EuropeGreater ChinaWestern Europe403020100501519252935394549555965+1519252935394549555965+1519252935394549555965+403020100501519252935394549555965+1519252935394549555965+1519252935394549555965+Age group1997 male2023 male1997 female 2023 fema

226、le Total frst wave regionsAdvanced AsiaNorth AmericaCentral and Eastern EuropeGreater ChinaWestern Europe403020100501519252935394549555965+1519252935394549555965+1519252935394549555965+403020100501519252935394549555965+1519252935394549555965+1519252935394549555965+20231997Age groupPDFWeb Exhibit of

227、How has labor intensity evolved?Weekly hours per capita over a life cycle,1997 and 202333Dependency and depopulation?Spain once again offers an interesting and extreme example of the impact of labor intensity on economic growth.Overall,weekly hours per capita there grew by an average 0.5 percent per

228、 year from 1997 to 2023,boosting GDP per capita growth by the same amount.The age mix shift did not add to or subtract from growth,meaning all additional hours worked were the result of increased labor intensity.But Spain,like other countries,not only will see a growing drag from the age mix,but may

229、 also struggle to sustain the very high growth in labor intensity of the recent past.Over the past quarter century,labor force participation among working-age females in Spain jumped from 60 to 81percent,an unrepeatable acceleration.Additionally,weekly hours per Spanish workerthose already in the wo

230、rkforcehave trended down for decades,and the country is currently discussing reducing the working week from 40 to 37.5 hours.21Productivity growth is the other lever underpinning GDP per capita growth,and generally the most important one.Across first wave economies,it has been the largest contributo

231、r to growth over the past quarter century(Exhibit 15).With demographics dragging down future GDP per capita growth,labor intensity,productivity,or the age mix will need to shift.By how much is the question we turn to next.Exhibit 1519972023202350EXAMPLE:Western EuropeBy regionBy countryAdvanced Asia

232、Central and Eastern EuropeGreater ChinaNorth AmericaAustraliaFranceGermanyItalyJapanSouth KoreaSpainUnited KingdomUnited States0.20.00.10.10.20.40.40.30.40.30.00.10.10.20.00.30.00.20.50.60.40.00.90.50.20.01.73.17.01.51.50.80.90.31.13.70.61.01.51.43.16.91.31.50.91.10.40.63.21.11.01.40.40.40.40.20.20.

233、20.40.70.40.70.80.10.2Age mix Age mixLaborintensityProductivityTotal GDP per capita growthProductivity growth was the main driver of growth in GDP per capita in the past quarter century.McKinsey&CompanyContribution of drivers to GDP per capita growth,19972023 and 202350,%Note:Figures may not sum to

234、100%,because of rounding.Source:ILOSTAT;World Population Prospects 2024,United Nations;The Conference Board;Oxford Economics;McKinsey Global Institute analysis0.30.40.400.81.034Dependency and depopulation?2.2.More workers,more hours,and more productivity neededA growing youth deficit has already set

235、 first wave regions on a trajectory for declining hours of work per person and thus slower GDP per capita growth through 2050 compared with the past quarter century.Maintaining past economic progress,let alone increasing it,will require measures to address the impact of demographic headwinds.In this

236、 section,we explore the three levers availablelabor intensity,productivity,and age mixto neutralize the drag on economic growth caused by shifting age mix.The third lever,influencing the age mix by increasing the number of working-age people,can be pulled via migration and higher fertility.However,t

237、he impact of higher fertility rates by 2050 would be negligible,as a baby born today would be barely joining the workforce.So we only analyze the impact of migration here.The conclusion is clear:pulling on only one of these three levers will be insufficient to achieve this goal in most countries,so

238、using some combination of all three will be needed to maintain growth and raise living standards(Exhibit 16).2.235Dependency and depopulation?To size the degree to which each lever must be pulled to sustain GDP per capita growth,we continue to rely on the all-else-being-equal approach we used for th

239、e analysis in the preceding section.We consider a target GDP per capita growth rate for each country equal to its annual growth over the past 25 years in all countries other than China and South Korea.In those countries,we lowered the target,since replicating their past growth would be an unrealisti

240、cally high goal given their much more advanced level of development today(see sidebar“Our approach to sizing the impact of demographic changes and how to counteract them”).22While the levers will likely need to be deployed in combination,each country can opt for a different“menu”of combinations,depe

241、nding on its characteristics,opportunities,and challenges.Exhibit 161.6x3x2.7x5.9x5.2x4.5x10 x1.6x3x2.7x5.9x5.2x4.5x1.7x1.5xn/a510 xORLabor intensity1How much each individual worksAge mix3Relative number of people per age group0319972023201223Needed to sustain past GDP per capita growthHOURSAddition

242、al hours worked per week per workerMIGRATIONAnnual fow as a%of 2023 populationPRODUCTIVITY GROWTHGrowth in GDP per hour worked,%PARTICIPATIONAdditional labor force participation rate(p.p.)Productivity2How productive each hour of work isChange in age mix,labor intensity,or productivity growth needed

243、to sustain GDP per capita growth in 202350Changing the age mix,labor intensity,or productivity alone cannot sustain growtha combination of all three is needed.Note:Beyond migration,increasing fertility rates can shift the age mix and have positive efects on GDP per capita in the long run,but 2050 is

244、 too early to see the impacts.1Assuming the same productivity growth as 19972023 continues.China and South Korea were less developed in 1997 and had very high productivity growth in the last quarter century;assuming constant productivity growth from the past would be unreasonable.For China,we assume

245、 productivity grows by 4.9%,in line with high-growth countries in the past,when they were at a similar development stage.For South Korea,we take the highest productivity growth across advanced economies in the last quarter century,2.0%.2Assuming the same hours per capita growth as 19972023 continues

246、.3Historical averages based on 20152019(prepandemic)annual averages.Where available,historical numbers are cited from national statistics agencies(Australia,France,Germany,Spain,United Kingdom).United Nations estimates are used for the remaining countries(China,Italy,Japan,South Korea,US).Future mig

247、ration required is calculated as the number of additional migrants required to maintain past growth if productivity growth remains constant.4While“maintaining past GDP per capita growth”is a reasonable target for highly developed economies,China and South Korea were not as developed in 1997.Therefor

248、e,their GDP per capita growth targets are reduced,based on the trajectories followed in the past by other countries when they were at similar stages of development.Chinas target is 4.9%per year,while South Koreas is 2.0%.5Historically,emigration has exceeded immigration in China,so needed migration

249、cannot be expressed as a multiple of past migration.Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisMcKinsey&CompanyAustraliaChina4FranceGermanyItalyJapanSouth Korea4SpainUnited Kingdom United States4.47.24.99.812.07.112.521.72.92.

250、7 2.5 5.6 2.7 5.1 8.3 3.97.4 13.6 1.5 1.51.57.80.80.90.31.13.70.61.01.51.75.51.01.40.92.03.31.41.11.81.36.60.30.70.30.92.30.30.50.8Historical annual migration(average of 201519)Annual migration needed if all migrants are workersAnnual migration needed if new migrants participate in work at same rate

251、 as past migrants36Dependency and depopulation?Both labor intensity and productivity must rise The first lever,increasing labor intensity,will be key.We analyze the overall increase in labor intensity needed to maintain past GDP per capita growth as well as what it would take if the full lift were a

252、ccomplished by just one of three population groups alone:males under 50 years of age,or“younger men”;females less than 50 years old,or“younger women”;and people older than 50years,or“older people.”Consider Germany,where GDP per capita grew an average of 1.1 percent per year over the past quarter cen

253、tury.Productivity growth over that period was 0.9 percent annually.Changes in the German age mix going forward are expected to reduce that growth by 0.4 percentage point.This means the country will need to fill a gap in GDP per capita of 0.6 percentage point per year if past productivity growth is t

254、o persist,which translates into an overall increase in labor intensity of 2.2more hours of work per capita per week,or 5.1 hours per worker.For any of the three cohortsyounger men,younger women,or older peopleto achieve that on their own,weekly hours per capita would need to expand by 12.7 hours,13.

255、2 hours,and 6.1 hours,respectively by 2050(Exhibit 17).Hours per capita can in turn be increased in two ways:more people workinghigher labor force participationor more hours worked per worker in the labor force.While the latter is a possibility,recent trends in hours worked have gone in the opposite

256、 direction:German workers today work Exhibit 172023 actual hours per capita2050 hours needed to maintain GDP per capita growth(holding other two groups hours constant)Additional hours per capita per week needed in Germany by 2050 to sustain GDP per capita growth,hours12.2Boosting labor participation

257、 by.Boosting hours per worker for existing workers by.2023 to 20502023 to 2050If the additional hours were worked only by.OROlder people aged 50796.118 p.p.17.945%63%27.044.9Seniors aged 507920.1(+6.1 vs today)Males aged 2049Females aged 2049Males aged 204912.742 p.p.17.190%100%236.753.8Seniors aged

258、 507944.6(+12.7 vs today)Males aged 2049Females aged 2049404040404040404040000Females aged 204913.254 p.p.19.681%100%229.549.1Seniors aged 507936.4(+13.2 vs today)Males aged 2049Females aged 2049ABAll peopleThis would be equivalent to.000 0McKinsey&CompanyExhibit of The required increases in hours w

259、orked to sustain GDP per capita growth are large:Germany example.1Calculated as the additional hours per capita per week required to ofset the 202350 age mix efect(if negative)and maintain 19972023 hours growth(if positive).This is equivalent to maintaining past GDP per capita growth if productivity

260、 growth remains constant.2Labor force participation among German men or women aged 2049 would require an increase that would drive labor force participation levels over 100%,which is impossible.Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institu

261、te analysis 37Dependency and depopulation?2.5hours less per week on average than 25 years ago,and the country is testing a four-day work week,which would further reduce labor intensity.23 Raising weekly hours worked by some groups,for example older people,may be feasible.Yet relying on only one coho

262、rt is challenging:if considering only existing workers,workers aged 50 to 79 would need to increase their weekly hours by 17.9 hours each,from an average of 27.0 hours today to 44.9 hours,to sustain past GDP per capita growth absent other changes.Germany could also increase labor force participation

263、,but the country already has one of the highest labor force participation rates in the world,at 90 percent among younger men and 81 percent of younger women.In fact,if the country were to rely solely on increasing labor force participation among these two cohorts to maintain past GDP per capita grow

264、th,younger workers labor force participation would need to exceed 100 percent,which is impossible.Germany could raise the participation of older people,which today is at 45 percent.However,the increase needed would be large,18 percentage points.The German example illustrates how increasing labor int

265、ensity growth,while necessary,is unlikely to do the job of maintaining GDP per capita growth on its own.Some countries,such as Australia,the United Kingdom,and the United States,have a somewhat easier task,given their more favorable demographic prospects(Exhibit 18).Others,like China,Italy,South Kor

266、ea,and Spain,have an even harder task than Germany.And keep in mind that this analysis depends on two important underlying assumptions.First,it assumes productivity growth continues to increase at the same rate,on average,Exhibit 18If worked only by.Additional hours per capita per week needed to sus

267、tain GDP per capita growth by 2050,hours1All peopleMales aged 2049Females aged 2049Older people aged 5079Total frst waveAustraliaChina2FranceGermanyItalyJapanSouth Korea2SpainUnited KingdomUnited States 2.21.2 2.7 1.1 2.2 2.7 1.8 3.4 4.7 0.7 0.7 12.56.1 15.2 6.2 12.7 16.7 9.1 24.4 29.1 3.8 3.7 13.66

268、.4 17.4 6.3 13.2 17.9 11.9 24.3 29.7 4.0 4.0 5.63.5 6.1 3.5 6.1 6.9 4.7 7.4 12.0 2.2 2.3 McKinsey&CompanyWeb Exhibit of Some countries need to increase hours worked more than others to sustain past GDP per capita growth.1Calculated as the additional hours per capita per week required to ofset the 20

269、2350 age mix efect(if negative)and maintain 19972023 hours growth(if positive).This is equivalent to maintaining past GDP per capita growth if productivity growth remains constant.China and South Korea were less developed in 1997 and had very high productivity growth in the last quarter century;assu

270、ming constant productivity growth from the past would be unreasonable.For China,we assume productivity grows by 4.9 percent,in line with high-growth countries in the past,when they were at a similar development stage.For South Korea,we take the highest productivity growth across advanced economies i

271、n the last quarter century,2.0%.2While“maintaining past GDP per capita growth”is a reasonable target for highly developed economies,China and South Korea were not as developed in 1997.Therefore,their GDP per capita growth targets are reduced,based on the trajectories followed in the past by other co

272、untries when they were at similar stages of development.Chinas target is 4.9%per year,while South Koreas is 2.0%.Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysis38Dependency and depopulation?as it did over the past 25 years.In a sc

273、enario with lower productivity growth,the growth in labor intensity needed would be even bigger,whereas if productivity growth increases,it would be smaller.Second,the target for each country is equal to its past GDP per capita growth,which in many cases is not high.Italys target,for example,is a me

274、re 0.4 percent annual growth.Should Italy want to achieve a healthier GDP per capita growth of,say,1.5 percent,similar to that of the United States or Australia,the growth in hours per capita required would jump from 2.7 to a whopping 7.9,assuming constant past productivity growth of 0.3 percent per

275、 year.Given the big,and in some cases unfeasible,requirements for increased labor intensity,countries also need to find ways to propel productivity growth,the second key lever.Assuming hours worked per capita grow at the same rate as in the past quarter century,productivity in most first wave countr

276、ies would need to grow between 1 and 2 percent a year to maintain past GDP per capita growth(Exhibit 19).That level of increase may seem modest,but in Germany,for example,it means doubling the past decades average rate of annual productivity growth of 0.7 percent.In Spain,productivity growth needs t

277、o increase by about four times,even assuming labor intensity grows at past rates.If labor intensity does not increasea plausible scenarioproductivity in Germany and Spain would need to grow by 1.5 percent and 1.9 percent,respectively,per year.Exhibit 19Exhibit of Productivity growth would have to ac

278、celerate substantially to match GDP per capita growth rates from 1997 to 2023.1Assuming the same hours per capita growth as 19972023.Calculated for frst 8 countries displayed as the percentage of annual growth in productivity needed to ofset the 202350 impact of aging(if negative)and maintain GDP pe

279、r capita growth matching 19972023(if positive).For China and South Korea,calculated as the productivity growth needed to achieve 4.9%and 2.0%annual GDP per capita growth,respectively.While“maintaining past GDP per capita growth”is a reasonable target for highly developed economies,China and South Ko

280、rea were not as developed in 1997.Therefore,their GDP per capita growth targets are reduced,based on the trajectories followed in the past by other countries when they were at similar stages of development.Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Gl

281、obal Institute analysisMcKinsey&CompanyItalyUnited StatesUnited KingdomFranceAustraliaSpainGermanyJapanChina2South Korea202.01.00.51.501.00.51.50681042Annual future productivity growth needed to sustain GDP per capita growth,1%1.71.41.41.11.05.53.30.91.82.02.2x3.0 x1.3x2.3x4.1x1.9x3.1x2.3x1.4x0.8xPr

282、oductivity growth,%1997201220122339Dependency and depopulation?While productivity can grow by raising capital investment and harnessing digital and automation technologies,most first wave countries have long struggled to do so.24 In fact,productivity growth has slowed in many of those countries over

283、 the past decade.For instance,US productivity over the past decade grew 0.8 percent per year on average,much less than its earlier annual growth of 2.0percent.From the beginning of 2023 through the second quarter of 2024,productivity growth spiked in the United States to top 2 percent per year,while

284、 it flatlined in many Western European countries and Australia.This suggests that the United States may be better positioned to jump-start growth,although that remains an open question.The future holds opportunities and risks for productivity growth everywhere.For example,while AI promises to propel

285、 productivity,increasingly fragmented global value chains and the growth of traditionally low-productivity service sectors like healthcare due to increasing longevity could restrain productivity growth.25China faces a special challenge.It is a first wave country because of its current demographic pr

286、ofile,but its GDP per capita of$21,000(after adjusting for purchasing power)is closer to that of later wave regions.The countrys population is aging faster than almost anywhere else on Earth due to its low and declining fertility rate.To achieve a 4.9 percent growth target,China would need to grow i

287、ts productivity by 5.5 percent a year,on average,through 2050 to counteract the demographic shift.This target is challenging,though not unattainable.While Chinese annual productivity growth over the past quarter century has been impressive,above 8 percent,it has slowed down more recently.Since the p

288、andemic and through 2023,Chinese productivity grew by 5.2 percent annually.26 As the country develops further,maintaining such very high rates of productivity growth will not be an easy feat.All in all,relying on either of these levers,labor intensity and productivity growth,to offset the impact of

289、the demographic shift on its own is unlikely to do the job.Fortunately,countries can use them in combination(Exhibit 20).The possible combinations of hours and productivity growth needed to maintain GDP per capita growth vary by country.For example,Germany could achieve past growth by increasing pro

290、ductivity at 0.9 percent per year while also increasing hours of work per capita by 2.2 or,alternatively,by growing productivity at 1.4 percent and hours of work per capita by 0.5hour.40Dependency and depopulation?It could also attain past growth with a middle point of productivity and hours between

291、 these two outcomes,for example productivity growth of 1.1 percent and an additional 1.6 weekly hours per capita(see sidebar“Isoquants help plot the many potential paths to sustaining growth”).What is clear is that most countries in the first wave will likely need to rely on both.Exhibit 20Weekly ho

292、urs per capita added by 2050056743223110531642ChinaSouthKoreaJapanItalySpainGermanyUnited StatesAustraliaFranceUnited KingdomGrowth in annualproductivity,202350,%Web Combinations of productivity and hours per capita growth to sustain GDP per capita growth1 Country maintains past productivity growth1

293、Country maintains past hours per capita growth1Note:For each economy,we show two points of the isoquants and the line that represents all possible combinations between the two.For visibility,we decide to not extend beyond the two points that mark the extremes of each isoquant,but the line technicall

294、y extends infnitely in both directions.1While“maintaining past GDP per capita growth”is a reasonable target for highly developed economies,China and South Korea were not as developed in 1997.Therefore,their GDP per capita growth targets are reduced,based on the trajectories followed in the past by o

295、ther countries when they were at similar stages of development.Chinas target is 4.9%per year,while South Koreas is 2.0%.Source:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisMany combinations of productivity and labor intensity growth co

296、uld sustain past GDP per capita growth.McKinsey&Company41Dependency and depopulation?Sidebar:Isoquants help plot the many potential paths to sustaining growthAn“isoquant”is an economic concept that describes different combinations of inputsin this case,productivity growth and increased labor intensi

297、tythat result in a given level of output;here,GDP per capitagrowth.Take Spain as an example(exhibit).Given the impact of the demographic shift there through 2050,if labor intensity and productivity growth stayed the same,GDP per capita would grow 0.2 percent per year on average,significantly less th

298、an the countrys average annual past GDP per capita growth of 1.1 percent.If Spain maintained its past growth in hours worked,or 1.5 extra hours of work per person per week,productivity would need to grow by 1.4 percent a year.By contrast,if productivity growth remained stable at 0.6percent a year,Sp

299、aniards would need to add 4.7 hours per week per capita to their current labor intensity levels.In fact,if Spain were unable to replicate the same labor intensity growth as in the past,which is not unlikely,the productivity growth needed to offset the demographic drag would exceed 1.4 percent and re

300、ach 1.9 percent.Exhibit1.1%/yr GDP per capita growth(sustaining past growth)387651042Weekly hours worked per capita added by 20500Growth in annualproductivity,202350,%1.61.41.00.60.22.01.81.20.80.4Due to the age mix drag,GDP per capita would grow just 0.2%per year if Spain had the same increase in p

301、roductivity and hours per capita as in the past.If productivity growth were the same as in the past0.6%per yearSpain would need to add 4.7 more hours per capita per week.If hours per capita growth was the same as in the past1.5 extra hours worked per capitaSpain would need annual productivity growth

302、 of 1.4%.332211Isoquant productivity and hours worked per capita growth to sustain past GDP per capita growth Country maintains past hours per capita growthPast growth in hours worked per capita0.5%/year Past productivity growth0.6%/year 0.2%/yr GDP per capita growth Country maintains past productiv

303、ity growthSource:World Population Prospects 2024,United Nations;ILOSTAT;The Conference Board;McKinsey Global Institute analysisIsoquants example:By increasing both hours worked per capita and productivity growth,Spain can sustain its past GDP per capita growth.McKinsey&Company42Dependency and depopu

304、lation?The age mix can be changed,but with insufficient impact by 2050The third lever to counteract the coming demographic drag is to influence the age mix itself,shifting the path already projected by the United Nations.This would require changes in fertility rates or migration patterns.While incre

305、asing fertility rates is critical for population growth over the long term,babies born today will barely have entered the labor market by 2050,reducing the potential impact of higher fertility rates over much of the next quarter century.27 Migration can more immediately help countries grow their wor

306、king-age population.However,the increase in migration needed to maintain GDP per capita growth is significant.For example,the United Kingdoms historical net migration is about 270,000 individuals per year,or 0.4 percent of its 2023 population annually,on average.28 Holding past productivity growth c

307、onstant,sustaining the countrys past GDP per capita growth rate would require an average additional 140,000 migrant workers annually assuming that all migrants work,which would mean increasing the countrys population by 0.6 percent per year through 2050.But not all migrants work,at least right away,

308、which increases the number of people of working age but does not necessarily boost GDP.Assuming that the share of new migrants to the United Kingdom participating in the workforce is the same as in the past,or 80 percent,then the annual flow of migrants needed to maintain GDP per capita growth in th

309、e United Kingdom at past levels would need to be an additional 185,000 above historical levels,or a total of 0.7 percent of the current population annually.29 Migrants also come with relatives who may not workfor example,childrenso the real number needed would be even larger.Other metrics illustrate

310、 the scale of migration required to maintain the economic status quo.For instance,new research estimates that if advanced economies relied on migration alone to maintain support ratios at todays levels,in many cases as much as half of their populations would be foreign born by 2050,assuming each mig

311、rant brings one dependent.3043Dependency and depopulation?2.3.A growing societal bill for dependencyOlder people consume more than younger people,directly through purchases and indirectly via in-kind transfers such as government-financed healthcare services,even as their income from working decrease

312、s and eventually disappears at retirement.This creates an imbalance between total consumption and labor income earned by seniors within a given year;we call this difference the senior gap.31This gap requires funding via public or private means.In first wave regions,the largest source of funding come

313、s from governments taxing workers over their careers and making transfers in the form of pensions and in-kind services like healthcare to seniors.Additionally,some older households finance part of their consumption by spending savings that they accumulated while working.As numbers of older people gr

314、ow,all else being equal,aggregate senior gaps will widen.To date,increasing public transfers,mostly pensions,and rising asset prices have funded growing consumption among seniors.Whether and for how long these sources are sustainable as life expectancy extends remains an open question.Pension system

315、s in first wave economies,especially those with defined-benefit programs,are already stretched.If returns on assets decline or public finances cant stretch any further,the only way to avoid savings depletion or falling consumption among seniors is some combination of the levers previously describedi

316、ncreased labor intensity and higher productivity growth over the short haul and effective migration and accelerated fertility rates in the longer term.Older people consume marginally more than younger peopleOur analysis of consumption encompasses out-of-pocket spending by households and individuals,

317、which we refer to as direct consumption,and,when data is available,in-kind services paid from the public purse.Such transfers are often an important share of consumption,especially in first wave economies with generous public education and healthcare programs.For example,the average German receives$

318、8,000 in public in-kind transfers a year,in addition to$30,000 they spend on direct consumption.32In most first wave regions,older individuals consume marginally more even as they earn less.33 In places where this isnt true,consumption doesnt fall substantially as people age.Both direct and in-kind

319、consumption grow with age.In 2023,those 65 years and older had the highest consumption per capita in most first wave regions(Exhibit 21).Compared with overall average direct consumption by all age groups,this cohort spent 16 percent more per capita annually in both Western Europe and Advanced Asia.W

320、hen measuring direct and in-kind consumption together,the discrepancy is even larger:seniors“spent”24 percent more per capita a year in Western Europe and 22 percent more in Advanced Asia.2.344Dependency and depopulation?The aggregate senior gap will grow Labor income and consumption change across a

321、 life cycle.Children are dependent on their families.When they reach adulthood,they join the workforce and generate income that may exceed their consumption needs,allowing them to save and accumulate wealth.At retirement,labor income decreases but consumption does not,creating a deficitthe senior ga

322、p.34 In the United States,when direct and in-kind consumption were combined,the gap equaled$60,000 on average per person 65years and older in 2023(Exhibit 22).Exhibit 21Advanced Asia01430441529456415294564152945641529456415294564Age group65North America014304465Western Europe014304465Greater China01

323、4304465Central and Eastern Europe0143044651Average per region calculated as unweighted average of index-based consumption per country,with base 100=average consumption per people in 2023 for each country.2Data on in-kind consumption,necessary to calculate total consumption,is not available for all c

324、ountries.In Advanced Asia,total consumption is available only for Australia,Japan,and South Korea;in North America,only for the United States;in Western Europe,only for Austria,Finland,France,Germany,Italy,Spain,and the United Kingdom;and in Greater China,only for Mainland China.No data on in-kind c

325、onsumption is available for countries in Central and Eastern Europe.Source:World Population Prospects 2024,United Nations;OECD;World Data Lab;McKinsey Global Institute analysis1008060140120Exhibit of Older people consume marginally more than younger cohorts.McKinsey&CompanyAverage consumption per pe

326、rson in 2023,1 indexed to 100=average consumption per capita in 2023Direct consumptionTotal consumption2(direct and in-kind)45Dependency and depopulation?As the share of older people rises,all else being equal,aggregate gaps will widen considerably by 2050.35 To compare the burden over time,we divid

327、e the aggregate gap by the total labor income of each region.Take North America,where it was 26 percent in 2023(Exhibit 23).This means that the total consumption of seniors not funded by their own labor income was equivalent to 26 percent of all labor income in the country.Seen another way,this is t

328、he equivalent of a flat tax that,if imposed on all worker incomes,would be required to cover the total senior gap.As the shift in age mix intensifies,North Americas gap will rise to 34 percent by 2050 and go from 33 percent to 46 percent in Western Europe and from 33 percent to 48 percent in Advance

329、d Asia.Exhibit 22Example:US labor income and consumption per capita by age group,2023,$,2021 PPPGap between labor income and consumption per capitaConsumptionLabor incomeNote:Consumption includes direct,out-of-pocket spending as well as the value of in-kind public goods and services.Source:National

330、Transfer Accounts Project;World Data Lab;McKinsey Global Institute analysisThe senior gap measures the diference between labor income and consumption among individuals 65 years and older.McKinsey&Company80,00070,00060,00050,00040,00030,00020,000010,000Younger than 15Senior gap$60,000$13,00015293044A

331、ge group,years456465 and older$34,000$22,000$23,00046Dependency and depopulation?The size of the gap may vary between countries based on consumer preferences.For example,consumers in one economy may choose to save more during their working lives and consume more in retirement,while those in another

332、may choose to save less and consume less in retirement.Regardless of the gaps absolute size,it will rise across all first wave economies.Individual seniors fund retirement in many ways:by saving directly during working years,by earning returns on assets,through public pensions,transfers and subsidiz

333、ed in-kind services,and through familial care and private transfers.In recent years in advanced economies,two of these sourcespublic pensions,mostly in the form of pay-as-you-go programs,and asset price appreciation,particularly in family homes,which has buoyed accumulated savingshave been central to funding seniors consumption,but they may be increasingly hard to maintain for future generations o