RL

Accounting for Intergenerational Transfers: An Overview Ronald Lee University of California, Berkeley Asia’s Dependency ...

3 downloads 653 Views 127KB Size
Accounting for Intergenerational Transfers: An Overview Ronald Lee University of California, Berkeley Asia’s Dependency Transition: Intergenerational Transfers, Economic Growth, and Public Policy Nihon University Population Research Institute Tokyo, Japan November 1-3, 2007

I. Transfers and Human nature • Humans are a social and altruistic species as we know from primatologists and anthropologists. – Hunter/gatherers lived in groups – Invested heavily in their children – Shared food, spreading risk and helping the needy

• Humans are also competitive and individualistic • Human nature is a continuing struggle between these tendencies.

1

In today’s economies • Competitive behavior is expressed in market exchanges. – Buying and selling – Borrowing and lending – Saving, investing

• Altruistic and socially oriented behaviors are expressed in transfers, gifts with no quid pro quo. – Intergenerational transfers (vertical) – Horizontal transfers (risk sharing, need based)

In today’s economies • Public sector transfers can be a very large share of the economy. – Not just a quaint carry-over from pre-capitalist days. – Deeply rooted in our human natures. – Should not be viewed as unnatural.

• However, extent of transfers to the elderly is new and population aging will drive their increase. • Resulting pressures on public budgets and families will lead to a major 21st century drama.

2

• National accounts cover market transactions but also public transfer programs. • National accounts do not attempt to cover private or familial transfers. • National accounts do not attempt to include the age dimension of flows or exchanges.

National Transfer Accounts (NTA) • Extend standard National Accounts in both ways – Add the age dimension – Add private transfers occurring within and between households • parents rear children. • Adults care for co-resident elderly parents • Bequests at death

3

II. Organization of the NTA project • Lead Institutions: East-West Center and CEDA, UCBerkeley • Asia Regional Office: Nihon University Population Research Institute • Funding – National Institute on Aging – United Nations Population Fund – Academic Frontier Project (Japan) – International Development Research Centre (Canada) – MacArthur Foundation – Others • www.ntaccounts.org

Research Teams for 23 Economies

4

The National Transfer Accounts project is a collaborative effort of East-West Center, Honolulu and Center for the Economics and Demography of Aging, University of California - Berkeley Lee, Ronald, Co-Director Mason, Andrew , Co-Director Auerbach, Alan Miller, Tim Lee, Sang-Hyop Donehower, Gretchen Ebenstein, Avi Wongkaren, Turro

Takayesu, Ann Boe, Carl Comelatto, Pablo Sumida, Comfort Schiff, Eric Stojanovic, Diana Langer, Ellen Chawla, Amonthep Pajaron, Marjorie Cinco

Japan Key Institutions: Nihon University Population Research Institute and the Statistics Bureau of Japan, Tokyo, Japan. Ogawa, Naohiro, Country Leader Matsukura, Rikiya Maliki Obayashi, Senichi Kondo, Makoto Fukui, Takehiro Ihara, Hajime Suzuki, Kosuke Akasaka, Katsuya Moriki, Yoshie Makabe, Naomi Ogawa, Maki

5

Australia Key Institution: Australia National University Jeromey Temple, Country Leader Brazil Turra, Cassio, Country Leader Lanza Queiroz, Bernardo Renteria, Elisenda Perez Chile Key Institution: United Nations Economic Commission for Latin America and the Carribean, Santiago, Chile Bravo, Jorge Mauricio Holz

China Key Institution: China Center for Economic Research, Beijing, China. Ling, Li, Country Leader Chen, Quilin Jiang, Yu Taiwan Key Institution: The Institute of Economics, Academia Sinica, Taipei, Taiwan. Tung, An-Chi, Country Leader Lai, Mun Sim (Nicole) Liu, Paul K.C. Andrew Mason

6

France Wolff, Francois-Charles, Country Leader Bommier, Antoine Thailand Key Institution: Economics Department, Thammasat University. Phananiramai, Mathana, Country Leader Chawla, Amonthep (Beet) Inthornon, Suntichai India Key Institution: Institute for Social and Economic Change, Bangalore Narayana, M.R., Country Leader Ladusingh, L. Mexico Key Institution: Consejo Nacional de Población Partida, Virgilio, Country Leader Mejía-Guevara, Iván

Indonesia Key Institution: Lembaga Demografi, University of Indonesia, Jakarta, Indonesia. Maliki, Country Leader Wiyono, Nur Hadi Nazara, Suahasil Chotib Philippines Key Institution: Philippine Institute for Development Studies. Racelis, Rachel H., Country Leader Salas, John Michael Ian S. Pajaron, Marjorie Cinco Sweden Key Institution: Institute for Future Studies, Stockholm, Sweden. Lindh, Thomas, Country Leader Johansson, Mats Forsell, Charlotte

7

Uruguay Bucheli, Marisa, Country Leader Furtado, Magdalena Rodrigo Ceni Cecilia Rodriguez South Korea An, Chong-Bum , Country Leader Chun, Young-Jun Lim, Byung-In Kim, Cheol-Hee Jeon, Seung-Hoon Gim, Eul-Sik Seok, Sang-Hun Kim, Jae-Ho

Austria Key Institution: Vienna Institute of Demography Fuernkranz-Prskawetz, Alexia, Country Leader Sambt, Joze Costa Rica Key Institution: CCP, Universidad de Costa Rica Rosero-Bixby, Luis, Country Leader Maria Paola Zuniga Slovenia Sambt, Joze, Country Leader Hungary Key Institution: TARKI Social Research Institute Gal, Robert Medgyesi, Marton Finland Key institutions: The Finnish Center for Pensions And the Finnish Pension Alliance Vanne, Reijo Gröhn, Jukka Vaittinen, Risto

8

United States Key Institution: Center for the Economics and Demography of Aging Lee, Ronald, Country Leader Miller, Tim Ebenstein, Avi Boe, Carl Comelatto, Pablo Donehower, Gretchen Schiff, Eric Langer, Ellen

Kenya Mwabu, Germano Nigeria Soyibo, Adedoyin

9

III. Flow identity • Basic accounting identity for flows for individuals at each age: Inflows=Outflows • Here is somewhat simplified version

Inflows: labor income + asset income + private transfers received + public transfers received + borrowing + sale of assets

10

Outflows consumption + public transfers made (tax payments) + private transfers made + purchase of assets + lending + payment of interest

Stocks can be derived from these flows • Capital stock • Credit • Transfer wealth (at each age, present value of expected future net transfers) • Each has a public and private dimension.

11

• convenient to focus on labor income and consumption – Strong biological component to productivity by age, e.g. children and elderly are less productive – Strong biological component to consumption by children

• Age schedules of labor income and consumption are centrally important for many theories – Samuelson’s consumption loan model – Diamond’s overlapping generations model – Life cycle savings theory

III. NTA estimates of age profiles of labor income • Fifteen countries (others are now available but not yet included). • Profiles are standardized relative to average labor income at ages 30-49 for each country. • In every case, our estimates are per member of the population, whether working or not, and averaged across sex at each age. • Arranged into three groups of five each by gdp per capita, adjusted for purchasing power parity (PPP).

12

Average Labor Income Profiles Grouped By GDP Per Capita 1.20

Labor Income (Scaled)

Rich have later peak then sharp drop and early retirement.

Highest 5 Middle 5 Lowest 5

1.00

0.80

0.60

0.40

And keep working longer

Poor start working younger

0.20

0.00 0

10

20

30

40

50

60

70

80

90

-0.20

Age

Two Contrasting Labor Income Age Profiles Reflect Policy Differences: Austria (2000) and Chile (1997) 1.20

Labor Income (scaled)

1.00

0.80

Chile

0.60

Austria 0.40

0.20

0.00

0

10

20

30

40

50

60

70

80

90

100

Age

13

What is going on here? Interpretations from Chile and Austria teams • Chile: Old age labor income may have been driven up by the new funded pension system with private accounts. • Austria: – The structure of the educational system and apprenticeship program encourages early work by teenagers. – The structure of the retirement system encourages very early retirement.

• So transfer systems help shape labor income

Average Consumption Profiles (scaled) Grouped By GDP Per Capita In rich countries the elderly consume publicly provided health care and long term care

1.60

Highest 5 Middle 5 Lowest 5

Consumption (Scaled )

1.40

1.20

1.00

In poor and middle countries with high coresidence of elderly, adult consumption is very flat across age.

0.80

In poor countries, investments in human capital are low

0.60

0.40

0.20

0.00

0

10

20

30

40

50

60

70

80

90

Age

14

Consumption Profiles (scaled) for 5 Middle Per Capita GDP Very high expenditures on education in some middle countries. Japan similar.

1.40

C o n s u m p t io n ( s c a le d )

1.20

1.00

0.80

0.60

Chile, 1997

Slovenia, 2004

South Korea, 2000

Taiwan, 2003

0.40

Uruguay, 1994

0.20

0.00 0

10

20

30

40

50

60

70

80

90

Age

Components of US Consumption, 2003 Private consumption is rising with age until early 60s, even before public health spending becomes important. Note decline in private health spending after age 65: Medicare.

Dollars (US, 2000)

40000

Public Health

Private Edu Public Edu

Private Health Private Durables

20000

Private Other

Public Other

0 0

10

20

30

40

50

60

70

80

90

Age

15

IV. Some theory—why these age profiles matter

The average ages at which income is earned and consumed • Calculate the average ages – Use the actual population age distributions – Multiply the age the age profiles – Find Ayl and Ac

• These average ages summarize a lot of information about the economic demography.

16

Consider the average age difference between consuming and earning, Ac – Ayl • Is income is spent before or after it is earned, on average? – If Ac – Ayl > 0 income is consumed after it is earned, and it must be held in some form of wealth between earning and consuming, so there is a demand for positive wealth. – If Ac – Ayl < 0 income is consumed before it is earned, so there is a demand for negative wealth, or credit. – If Ac – Ayl = 0 there is no net demand for wealth.

Life cycle wealth W • This amount of wealth is required to achieve the observed age profiles of consumption and labor income. • Life cycle wealth is proportional to the gap in average ages (Willis, 1988): • W=(Ac – Ayl)c Where c is per capita consumption

17

Life cycle wealth can be capital, K, or transfer wealth, T. • At age x, T(x) is the Present Value of expected future (transfers received – transfers given) after age x. • T is the population weighted average of T(x). • Transfer wealth is the sum of public and familial: T = TG+TF • Children receive transfers first, and then make them later in life to their own children. – Their transfer wealth is negative.

• Transfers to elderly are made first, received later. – Their transfer wealth is positive.

• Overall, T is positive or negative depending on whether transfers to old or young dominate.

• In a closed economy: W = T + K • This is a fundamental accounting identity. • Alternatively, K = W - T. – The bigger are transfers to the elderly, the less capital there will be. – The bigger are transfers to children, the more capital there will be.

18

Fertility, population growth and aging • When fertility is higher, – population growth is more rapid – Population is younger.

• When fertility is lower – Population growth is slower or negative – Population is older – Japan is champion!

Effects of population growth depend on transfer wealth T • Consider the present value of consumption over the whole lifetime, C. • The effect of a change in fertility, or population growth rate, on C is proportional to transfer wealth T (Willis, 1988)

• d ln(C)/dn = T/c

19

• Higher fertility, faster pop gr, and younger pop – Raise life time consumption C if net transfers are upward to the elderly (T>0) – Reduce life time consumption C if net transfers are downward to children (T<0). – Have no effect if net transfers are 0 (T=0).

• This is a striking, non-obvious result • It depends not just on govt programs, but on whole age pattern of earning and consuming, and how periods of dependency are funded.

Conclude that patterns of transfers are very important • Insights carry over to more general cases, beyond golden rule. • Much depends on whether net transfers are upward to elderly or downward to children. • The NTA project has carried out a number of simulations and dynamic optimization exercises that confirm the importance of transfers in a more realistic circumstances.

20

Summarizing the direction of reallocation of income (in golden rule steady states) •

The population-weighted average age of labor income tells us the average age at which a dollar of income is earned, Ayl. Similarly for consumption, Ac. The difference, Ac – Ayl, tells us whether on average income is spent before or after it is earned.

• • –

– –

If Ac – Ayl > 0 income is consumed after it is earned, and it must be held in some form of wealth between earning and consuming, so there is a demand for positive wealth. If Ac – Ayl < 0 income is consumed before it is earned, so there is a demand for negative wealth, or credit. If Ac – Ayl = 0 there is no net demand for wealth.

• This amount of wealth is required to achieve the observed age profiles of consumption and labor income. • It is called “life cycle wealth” and denoted W • An important identity (Willis, 1988): W=(Ac – Ayl)c Where c is per capita consumption.

• Life cycle wealth is proportional to the gap in average ages.

21

Wealth does not have to be assets • Life cycle wealth can be either capital or transfer wealth in closed economy. • Transfer Wealth, T. – At age x, T(x) is the Present Value of expected future (transfers received – transfers given) after age x. – T is the population weighted average of T(x). – Transfer wealth is the sum of public and familial: T = TG+TF

• Wealth can also be held as credit, but in a closed economy credits and debts must sum to zero.

• In a closed economy: W = T + K • This is a fundamental accounting identity.

Transfer wealth and effects of pop growth and aging • Let C be the present value of life time consumption. • Let dn be a small change in pop growth rate due to a difference in fertility. • Willis (1988) showed : – d ln(C)/dn = Ac – Ayl – K/c

or substituting,

– d ln(C)/dn = T/c • Says higher fertility, faster pop gr, and younger pop – Raise life time consumption if net transfers are upward to the elderly – Reduce life time consumption if net transfers are downward to children. – Have no effect if net transfers are 0.

22

In case of T=0, no net transfers • All wealth is held as capital, like pure life cycle savings. • Changing population age distribution causes just enough change in savings to offset the change in population growth rates.

V. NTA data on average ages of consuming and earning, Ac and Ayl

23

• NTA data will let us evaluate T = TG + TF but not yet. • Now look at the direction reallocation flows, Ac – Ayl, through all means, including transfers. • First look at averages for the same three income groups as before.

Average Consumption-Earning Gap by average income groups: 15 NTA countries 0 -1

0

10000

Average Ac-Ayl

30000

40000

In rich countries, the average dollar moves only very slightly downward.

-2 -3 -4 -5

In poor countries, the average “dollar” travels downward 8 years to younger ages.

-6 -7 -8 -9

20000

Low Income

Middle Income

High Income

Average income by group

24

Avg Age of Av Age of Cons - Av Age of Labor Inc

Average Consumption-Earning Gap by GDP 4 Austria, 2000

2 Slovenia, 2004

0

Japan, 2004 France, 2001 Sweden, 2003

Uruguay, 1994

-2

US, 2003

-4 Costa Rica, 2004 S. Korea, 2000Taiwan, 2003 Thailand, 2004

-6

Chile, 1997

-8 Indonesia, 1999 India, 1999

-10 Philippines, 1999

-12 0

5000

10000

15000

20000

25000

30000

35000

40000

PPP-Adjusted GDP per Capita

Why does the age gap vary across these income groups? • First look at average ages using the same population age distribution for all countries, so result depends only on the shapes of the age profiles.

25

Average Ac-Ayl weighted by actual pop

Average Consumption-Earning Gap by average income, using fixed pop weights 0 -0.2 0 -0.4 -0.6 -0.8 -1 -1.2

5000

10000

15000

20000

25000

30000

35000

Here we see the same pattern: in poorer countries resources travel downwards farther. But the difference here is tiny, only .5 years, compared to a total difference of 7 years before.

-1.4 -1.6 Per capita gdp, ppp adjusted

V. Population Aging, Consumption Patterns and the Age Gap • Next look at the age gap in relation to the average age of the population

26

Average Consumption-Earning Gap by Average Age of Population 4

Austria, 2000

2

Slovenia, 2004 0

Japan, 2004 Uruguay, 1994 US, 2003

Av Age Gap

-2

France, 2001 Sweden, 2003

-4

S. Korea, 2000 Costa Rica, 2004 Thailand, 2004Taiwan, 2003

-6

-8

Chile, 1997 Indonesia, 1999 India, 1999 -10

Philippines, 1999 -12

20

25

30 35 Ave Age of Population

40

45

Demographic change • Income, population aging, and rising consumption age are all connected. • Population aging causes big changes in these average ages and in life cycle wealth, W. • But aging is also caused by changes in transfers, perhaps.

27

Ln Total Human Capital Spending per Child vs. Fertility (Health and Education only, up to 18 and 26, respectively)

ln(Total HK)

2.00 1.50 1.00 0.50 0.00 0.00

2.00

4.00

Fertility (TFR)

The quantity-quality tradeoff in fertility • Causal story behind chart is not clear. • Desire to invest more per child may cause fertility decline. • Desire to have fewer children may enable increased investments in each. • Virtually the entire relationship is due the relation between the TFR and public education costs per child – but this still does not tell us the causal direction.

28

Public pension systems may also help cause fertility decline • When old age support comes from family, there is a strong motive to have enough children. • When old age support comes from public pensions, then this motive is gone. • Same applies to other public transfers that may substitute for children’s contributions: health care, long term care, e.g.

VII. Will the rising costs of supporting the elderly crowd out investment in children?

29

Ratio of Total Consumption by Old to Children versus Average Age of Population Ratio of Agg Elder Cons to Agg Child Cons

1.40 Japan

1.20 1.00 0.80 US 0.60 0.40 0.20

Philippines 0.00 20.0 25.0 30.0 35.0 40.0 Average Age of Population

45.0

• The ratio of aggregate spending on the elderly to spending on children is bigger by a factor of ten in Japan than in Philippines or Indonesia. • Now look at spending per child vs per elderly.

30

Per Capita Consumption of Young and Old Relative to Working Ages More for Elderly 1.3 US, 2003 Philippines, 1999 1.2 Uruguay, 1994

Sweden, 2003 Japan, 2004

Ave C of Ages 65+

1.1

Slovenia, 2004 India, 1999 France, 2001 Costa Rica, 2004 Austria, 2000

1.0

Thailand, 2004 Chile, 1997

0.9

Taiwan, 2003 S. Korea, 2000

Indonesia, 1999 0.8

More for Children

0.7

45º 0.6 0.6

0.7

0.8

0.9 1.0 Ave C of Ages 0-20

1.1

1.2

1.3

Per Capita Consumption of Young and Old Relative to Working Ages More for Elderly Strong spending on elderly US, 2003

1.3

Philippines, 1999

1.2 Uruguay, 1994

Sweden, 2003 Japan, 2004

Ave C of Ages 65+

1.1

Slovenia, 2004 India, 1999

Strong investment in children

France, 2001 Costa Rica, 2004 Austria, 2000

1.0

Thailand, 2004 Chile, 1997

0.9

Indonesia, 1999

Taiwan, 2003 S. Korea, 2000

0.8

More for Children

0.7

45º 0.6 0.6

0.7

0.8

0.9 1.0 Ave C of Ages 0-20

1.1

1.2

1.3

31

Ratio of Per Capita Elder Cons to Child Cons

Ratio of Per Capita Consumption of Elderly to Children by Average Age of Population 2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 0.20 0.00 20.0

US Phi li ppi nes Japan

Kor ea

Tai wan

Sl oveni a

No evidence here that population aging is crowding out spending per child. 25.0

30.0

35.0

40.0

45.0

A v e ra ge A ge o f P o pula t io n

• Now Japan is in the middle • Philippines has relatively high consumption per elderly • Taiwan, Korea, and Slovenia again stand out for high investment in children. • Crowding out is unclear because: – Population aging is due mainly to low fertility – Low fertility is associated with greatly increased human capital investment per child

32

VIII. Substitution of public and private transfers to elderly • Note that elderly also get income from – Continuing labor – Assets

• Cross nationally, do countries that have larger public transfers to elderly also have smaller private ones? • Preliminary data for only 9 countries

Average Private versus Public Transfers to the Elderly per capita (65+) 0.6 Avg Annl Public Transfer (Scaled)

Chile, 1997

0.5

US, 2003

Costa Rica, 2004

0.4 Uruguay, 1994

0.3 0.2 Taiwan, 2003

0.1

South Korea, 2000 Indonesia, 1999

0

Philippines, 1999

Thailand, 2004

-0.1 -0.3

-0.2

-0.1

0

0.1

0.2

0.3

Avg Annl Private Transfer (Scaled )

33

Average Private versus Public Transfers to the Elderly per capita (65+) 0.6 Avg Annl Public Transfer (Scaled)

Chile, 1997

0.5

US, 2003

Costa Rica, 2004

0.4 Uruguay, 1994

Large private and small public

0.3 0.2 Taiwan, 2003

0.1

South Korea, 2000 Indonesia, 1999

0

Philippines, 1999

Thailand, 2004

-0.1 -0.3

-0.2

-0.1

0

0.1

0.2

0.3

Avg Annl Private Transfer (Scaled )

Average Private versus Public Transfers to the Elderly per capita (65+) Large public and small private

Avg Annl Public Transfer (Scaled)

0.6

Chile, 1997

0.5

US, 2003

Costa Rica, 2004

0.4 Uruguay, 1994

Large private and small public

0.3 0.2 Taiwan, 2003

0.1

South Korea, 2000 Indonesia, 1999

0

Philippines, 1999

Thailand, 2004

-0.1 -0.3

-0.2

-0.1

0

0.1

0.2

0.3

Avg Annl Private Transfer (Scaled )

34

• These topics are just a sampling of some preliminary comparisons • Now consider broader uses of NTA.

VIII. What is the relevance of age patterns of transfers, labor, consumption, and population aging? • Development and growth – Dependency and the First Dividend – Demand for wealth and the Second Dividend

• Efficiency – Do transfers displace capital? – Do transfer systems distort labor supply decisions?

• Investment in children – Competition between children and the elderly for transfer resources? – Are we making net life time transfers (total bequest) to our children taking everything into account such as national debt and future transfer obligations to the elderly? Or are we in effect borrowing from them?

• Public Sector – – – –

Sustainability of Public Program Structures Intergenerational equity when demography or systems change Horizontal equity: public transfers by social class Design of public programs.

35

What is the role of NTA in addressing these questions? • These topics are difficult to study within a country – not much variation – interpretation of changes over time is difficult

• With 23 very different countries we have the variation to do crosscountry analyses. • NTA will provide comprehensible and comparable measures of the relevant quantities for the first time. • NTA age profiles for the public sector are inputs for analyses like – – – –

generational accounting Long term budget projections Simulation studies of intergenerational equity under different policies Historical trends in generational accounts

• NTA age profiles are inputs for simulations of saving, capital accumulation, economic growth and development. • NTA can also describe transfer patterns by education, ethnicity, or sex.

Exciting times for the NTA project. • We are beginning to get comparable high quality data for many countries. • There has been strong interest in NTA results by different international agencies. • We are finding some surprisingly strong differences in patterns of labor supply, consumption, transfers, and capital accumulation across countries. • Researchers are making new discoveries about their own countries. • We will be hearing about some of this work over the next few days, and I greatly look forward to it.

36

END

37