- Email: [email protected]
Introduction: Special issue on exploring the generational economy
The Journal of the Economics of Ageing 5 (2015) 1–6
Contents lists available at ScienceDirect
The Journal of the Economics of Ageing journal homepage: www.elsevier.com/locate/jeoa
Introduction: Special issue on exploring the generational economy
The interplay between demography and economics is a rich and important area of interdisciplinary study involving researchers from economics, demography, sociology, political science, anthropology, psychology, and other fields of study. In the past, economic analysis incorporated demography using highly stylized models. Many studies and even the System of National Accounts takes the household as the basic decision-making unit treating the household as a black box. This obscures many important aspects of economic behavior, such as, the roles and economic effects of children, the elderly, and gender. In the past, macroeconomic models ignored population age structure or relied on highly simplified models of age structure. Population and labor force were treated as equivalent or populations were assumed to consist of only two or three adult age groups. These highly simplified representations of the world have been very valuable and yielded important insights about intergenerational transfers, credit markets, economic growth, and other macroeconomic processes, but they have obvious limitations. Research on the interplay between demography and economics has shifted over time to using more realistic models of the generational economy. In part, this is a natural outgrowth of the increase in computer power available on researchers’ desks (or in their laps). In part, it reflects the need to provide richer accounts of the demographic transition being experienced in almost every developing and developed country. Since the 1970s Family Economics has extensively explored decision-making within the family about childbearing, marriage, and divorce. The role of altruism and other motives for intergenerational transfers have been extensively explored. Models have also bridged the gap between the micro and macro levels of analysis, leading to a body of work sometimes called ‘‘family macro’’. The demographic transition has led to dramatic improvement in life expectancy that is an outcome, in part, of economic change but also represents an increase in welfare that is poorly captured by many economic models and economic statistics. The demographic transition has led to sharp changes in population age structure with more to come that are having a pervasive influence on our economies. The papers in this special issue were selected from among those presented at the Ninth Global Meeting of the Working Group on Macroeconomic Aspects of Intergenerational Transfers hosted by the University of Barcelona and held at Faculty of Economics and Business, 3–8 June 2013. The conference was attended by members of the National Transfer Account network and other researchers working to improve our understanding of the connections between economy and demography.
http://dx.doi.org/10.1016/j.jeoa.2015.04.001 2212-828X/Ó 2015 Elsevier B.V. All rights reserved.
An overview of National Transfer Accounts National Transfer Accounts (NTA) provides a conceptual framework and a system of accounts for the generational economy (Lee and Mason, 2011). The generational economy refers to the social institutions and economic mechanisms used by generations or age groups to produce, consume, share and save resources. Understanding the generational economy is important due to a fundamental feature of all societies – the economic lifecycle. Over an extended period at the beginning of life in all societies, children consume more than they produce through their labor – a gap we refer to as a life cycle deficit. This is followed by a lifecycle surplus over many years during which humans produce more through their labor than they consume. Over the final phase of the economic lifecycle in all contemporary societies for which we have estimates (other than some hunter–gatherer and subsistence swidden groups), humans again experience a lifecycle deficit consuming more than they produce through their labor. The significance of the economic lifecycle depends on the economic mechanisms and institutions that enable its existence. A very substantial portion of every nation’s resources must be reallocated across age – from the lifecycle surplus ages to the lifecycle deficit ages. The needs of children and youth are met through transfers – private transfers from their families and public transfers through government with the relative mix varying across countries. To a limited extent, young adults may rely on credit markets (student loans, consumer debt). The picture is very different for older adults. The relative importance of public and private transfers to the elderly varies greatly from country to country. Moreover, in many countries the elderly rely heavily on assets (asset income or dis-saving) to fund their lifecycle deficits. NTA has been developed to complement National Accounts by adding an age dimension, and in most respects is consistent with National Accounts. The focus on the individual in NTA rather than the household, however, leads to a much richer description of public and private transfers than in National Accounts. NTA provides estimates of transfers by purpose and by the age of the provider and the recipient. Estimates of both inter- and intra-household transfers are constructed. Moreover, NTA provides a consistent set of data on how asset income, debt, and assets vary over the lifecycle. NTA are constructed using a variety of sources of data from National Accounts, administrative records that document taxes and public spending, household surveys that contain information about income, consumption, and demographic characteristics, labor force surveys, and population censuses. More recently, time
2
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
use surveys are used in many countries to construct estimates of home production activities including child and elder care time. The methods for constructing basic accounts are described in the NTA Manual recently published by the United Nations Population Division United Nations Population Division (2013). Research teams in over 50 countries have constructed or are currently constructing NTA, sometimes for multiple years. Estimates for at least some components of basic NTA are available for 40 countries on the NTA website (Lee and Mason, 2014). The initial goal of the NTA network was to construct and analyze accounts for a recent year in countries that varied greatly in their political and economic systems, level of development, quality of data, and so forth. This collaborative effort allowed us to evaluate and to refine NTA methods. It also proved to be very useful for improving our understanding of the generational economy and the macroeconomic implications of changing population age structure. The contribution by Oosthuizen on the demographic dividend in South Africa is an excellent example of this kind of analysis. Many of the papers in this volume employ NTA data for multiple years to explore how features of the generational economy are evolving over time as populations age and as economies develop or respond to economic crisis. This is an important development and will lead to many research opportunities for exploring synthetic cohort data. Vogt and Kluge use NTA time series data of this sort to look at the role of West–East transfers in the convergence in life expectancy after reunification. The Patxot et al. paper decomposes changes in NTA across time, with the purpose of testing the sustainability of the per capita levels of Spanish consumption before the 2007 international financial crisis in face of population ageing. The paper by Racelis, Abrigo, and Salas compares the NTA estimates for 1999 and 2007 for the Philippines exploring also the role of migrant’s remittances, quite relevant in this country. The paper by Lai and Tung takes a longer perspective by investigating changes in the structure of transfers from 1985 to 2005. Finally, Albis et al. take a longer term perspective to analyze the evolution of the French lifecycle deficit for more than two decades (1979– 2005). The value of NTA can be enhanced through other extensions that are explored here. One active areas of research is to explore how the lifecycle and the age reallocation system varies by socioeconomic groups. NTA values in the basic accounts are aggregates or per capita values (averages). This is ideal for some applications, but for other applications it is important to know how lifecycle issues, intergenerational transfers, systems of support, etc. influence low- or middle-income groups. Papers by Mejia-Guevara and by Jimenez-Fontana introduce disaggregation by socioeconomic status (mainly measured by education) for Mexico and Costa Rica, respectively. Another important area of work is to use a broader measure of economic flows incorporating the value of goods and services that are produced primarily at home. Time transfers are a very important part of the intergenerational support system. Mothers, in particular, provide valuable services through childrearing, caregiving, and other activities that occur outside the market place. Estimating National Time Transfer Accounts (NTTA) provides a more realistic view of the costs of children, the economic contributions of women, and the costs and contributions of the elderly. The development of NTTA has been essential to another NTA-related enterprise – estimating accounts separately by gender. The papers by Zannella, Jimenez-Fontana, and Gal et al. illustrate this approach for the cases of Italy, Costa Rica and Hungary, respectively. Many of the papers in this special issue explore the implications of incorporating a more complete assessment of the economic value of work and the intergenerational transfers that arise primarily between family members. Hammer et al. provides a comparative analysis on the economic dependency ratio in eight
European countries using both the NTA profiles and a first approximation of the NTTA profiles. Going deeply into the role of transfers, the paper by Gal, Szbo, and Vargha uses NTA and NTTA profiles from Hungary to show the extent to which the cost of the elderly and of the children are socialized. The authors extend the work previously done in other countries by Lee and Donehower (2011) and Patxot and Rentería (2012) using NTA profiles, to include the time transfers (NTTA) in the picture.1 This reinforces the result that public transfers to children are far less important than public transfers to the old. This phenomenon, which occurs in other countries, is quite surprising if one takes into account that transfers to the old given to finance long-term care, pensions and health care, are by far more substitutable in the market than transfers to children. The first two papers in the special issue, authored by Tobias Vogt and Fanny Kluge and by Morne Oosthuizen, were selected as the outstanding papers presented at the conference, a prize sponsored by the Journal of the Economics of Aging and Elsevier. The paper by Vogt and Kluge ‘‘Can public spending reduce mortality disparities? Findings from East Germany after Reunification’’ investigates the impact of public spending on mortality disparities. The paper mainly focuses on the German case. The reunification of East and West Germany in 1990 and the improved conditions in the East thereafter provide a natural experiment that is used in this paper to evaluate the effect on mortality at older ages of increased pension benefits and increased health care spending. Before reunification, the gap between life expectancy in the East and West was large and growing. After reunification, life expectancy at 65 in the East increased rapidly and converged with that in the West, in contrast to Czechoslovakia, Hungary and Poland where mortality conditions are still similar to those in East Germany in 1990. This suggests that greatly increased financial transfers and spending on health care following reunification with West Germany played an important role. To investigate this idea more formally, a difference-in-differences analysis is used, finding that a one euro expenditure on health or pensions for older people in the East raised the length of life by 3 h. In further analysis, equations are estimated separately in East and West for the log of age and gender specific mortality by cause, where the covariates include average health care and pension expenditures by age, sex, region and period. The results show that expenditures on both pensions and health care reduce mortality, but health care is most important. Furthermore, expenditures in the East led to greater improvements than similar expenditures in the West. Morne Oosthuizen, in his paper ‘‘Bonus or Mirage? South Africa’s Demographic Dividend’’, explores a very important issue throughout the developing world – how best to achieve more rapid economic growth by harnessing the demographic dividend. The South African experience is important in many respects. Its fertility has declined further than elsewhere in the region and South Africa has had to deal with many serious problems arising as a consequence of apartheid. Oosthuizen addresses the first demographic dividend showing how the support ratio in South Africa is influenced by the particular features of labor income and consumption. Realizing the first dividend in South Africa has been a challenge because labor income is so low among young adults, a feature that has been found in some other African countries as well. Consumption patterns have also had an important influence on the support ratio. As Oosthuizen observes, however, it is unclear what policies might be pursued to influence consumption patterns in contrast to how low youth productivity might be addressed. Oosthuizen goes on to explore the second demographic dividend – how demographic change can be translated into higher
1
Patxot et al. (2012) discuss the related economic literature.
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
capital accumulation and investment in human capital. Spending on human capital in South Africa appears to reflect the quantity– quality tradeoff. Average public and private health and education spending per child is very similar to the levels found in other countries with similar rates of childbearing and similar levels of development. An important issue addressed by Oosthuizen is that human capital spending in South Africa is ineffective because of its failure to address the great disparities that persist. If this problem is not addressed, South Africa may fail to capture one of the most powerful pro-development channels of the second demographic dividend. The next three papers employ extensions of NTA data to document and to analyze important features of the generational economy in different contexts. Mejia Guevara, in his paper ‘‘Economic Inequality and Intergenerational Transfers: Evidence from Mexico’’, extends the standard NTA analysis by introducing socioeconomic status and by comparing NTA profiles at two points in time 1994 and 2004. The main issue addressed by this paper is the substantial decline in inequality over this period. The expansion of transfers, in general, and the conditional cash transfer program Progresa– Oportunidades, contributed in many ways including the expansion of basic education. The paper is interesting from a methodological perspective as the author disaggregates NTA estimates by socioeconomic status, using educational attainment as a proxy, in order to investigate if socioeconomic status influences age reallocations. As in the paper by Jimenez-Fontana, the decomposition of NTA profiles by education status, allows us to see differential patterns in the country, some of which resemble those observed in developed countries. For example, the most educated among the elderly show consumption increasing with age, partly due to the increase in public consumption, as in some developed countries, while labor income is not yet reduced by early retirement. The insights of the author turn out to be very instructive when looking at the age reallocations at different education levels. In brief, the transfer system seems to be acting as redistributive tool that operates most completely to the benefit of the less educated. ‘‘On the one hand, while for the less educated the main financing source of child consumption is public transfers (48.0), followed by private transfers (39.8%) and labor income (11.9%); private transfers are the main financing source for the most educated (96.7%). On the other hand, regarding consumption of the elderly in the lowest SES group, consumption is funded by asset-based reallocations (36.4%), public transfers (37.9%) and labor income (33.2%), while it is funded entirely by asset-based reallocations (108.2%) in the highest socioeconomic group’’. Despite the development of asset markets, intrafamily arrangements still play a role in Mexico. Quite interestingly, the elderly in the upper SES group rely on their assets, and transfer a substantial amount of resources to their children and grandchildren in the form of private transfers. This crowding out effect is observed in all socioeconomic groups, but the magnitude is different, ranging from 7.5% in the less educated to 44.2% for the most educated. The magnitudes have also increased from 1994 to 2004 as a result of the increase in transfers during the period. The paper ‘‘Reallocation of Resources between Generation and Genders in the Market and Non-Market Economy. The Case of Italy’’ by Marina Zannella provides a complete picture of intergenerational transfers in Italy combining both monetary transfers, both market and non-market transfers, with time transfers. The author begins by pointing out that investing in human capital of children and caring for the elderly are the two major tasks for intergenerational transfers. In Northern Europe the public sector plays a major or dominant role in both whereas in Southern Europe the family plays a greater role, particularly for children.
3
This paper examines the system of intergenerational reallocation of resources in Italy in 2008, including both market goods as in National Transfer Accounts, NTA, and also transfers of time as in the new National Time Transfer Accounts, NTTA. This permits a thorough examination of gender differences in production and consumption, since in Italy men are more active in market work while women, who have relatively low labor force participation, are more active in home production and care. This paper presents many interesting results. Starting with standard NTA by gender, the authors show that males and females consume very similar amounts at each age, except that after age 70 men begin to consume considerably more than women. Men have far higher labor income than women, roughly twice as great at each age. Consequently, while male and female children benefit equally from family transfers until age 18 or so, after this the net family transfers received by women are much greater than those by men, and indeed after age 25 men make substantial net familial transfers at every age whereas women continue to be net recipients at every age except in their forties when they make small net transfers to others. This is a misleading picture of women’s economic activity, however, since so much of female work is non-market activity in the home. The NTTA based on time use data shows that women produce roughly twice as much in home production after age 20, and girls begin producing more than boys even before they are ten years old. Transfers of home production are the mirror image of transfers of market goods, with females making large net transfers to males after age 20 while males are net recipients throughout their whole lives. In NTTA, children are much more costly and have a much larger life cycle deficit, than in NTA, due to their time costs. At other ages, the life cycle deficit for sexes combined appears very similar with and without home production included (NTTA vs NTA) because the ages of women providing home production and the men consuming it are similar, so it tends to average out. From birth through the mid-twenties, but particularly before age 5, familial transfers are much larger than public ones. Zanella suggests that Italy should provide greater public support for children and young adults, reducing the familial role in this regard. This would aid gender balance since the familial role is in fact largely female. Pamela Jimenez-Fontana, in ‘‘Analysis of Non-remunerated Production in Costa Rica’’, focuses on non-market production in Costa Rica which she estimates at 40% of GDP. The author presents NTA and NTTA profiles for males and females, finding heavy specialization by gender. In particular she finds that there is strong specialization by type of production, so that men specialize in the labor market and women specialize in home production. Hence, the analysis by gender implies that taking into account home production reduces the gender gap by half in Costa Rica. She also finds that there is specialization in the different tasks in home production. Women specialize in the production of care and other activities like cleaning, cooking, and laundry, while men specialize in car and farm maintenance and spend some time to travel to and from work. Overall, the activities that men specialize in represent a small proportion of all non-remunerated production. This situation is not surprising in a developing country like Costa Rica. The author is interested in the future evolution of labor division within the household, including help received from the rest of the family network. Hence, the author analyses to what extent gender specialization is motivated by a higher opportunity cost related to educational attainment. To that end, the NTTA profiles are disaggregated by education level. The results show that education matters: Women with high educational level dedicate less time to activities like cooking and cleaning, which may be explained by the high opportunity cost of staying at home doing these duties without economic compensation. But still, they spend
4
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
more time on these tasks than men. Regarding the production of care, there is also heterogeneity for both men’s and women’s profiles by educational level. For women, one can observe a first peak in the age profile of the production of care, which could be identified as taking care of children. Women with low levels of education have two initial peaks and women with completed high school or more only have one. These peaks are coherent with the observed TFR in Costa Rica by educational level. The production of care performed by women suggests a postponement of fertility’ and male’s profiles mainly follow this trend; women with high levels of education have the peaks of production of care later in life. Quite interestingly, the reduction of time spent in cooking and cleaning activities by females is compensated by an increase in time engaged in household management. Educated females, a growing group in Costa Rica, spend more time on these kinds of tasks, than women with primary education or less. In some sense, this trend represents women’s empowerment in the household and in the economy. Less educated women depend economically on their partners and family members. The next four papers focus on longitudinal issues and provide examples of how a future longitudinal NTA database can contribute to shed light on the role of intergenerational transfers. The first of these papers by Patxot, Renteria and Souto explores an important question in Spain and is described well by the title of their paper: ‘‘Can we keep the pre-crisis living standards? An analysis based on NTA profiles’’. Between 2000 and 2008, the period of analysis, demographic conditions were actually exceptionally favorable, due to the high proportion in the working ages and the low numbers of children and elderly. Using NTA data, the economic support ratio is shown to be at its peak around 2008, a situation which will be followed by decades of decline as the population ages. Furthermore, simulations show that demographic change from 2000 to 2008 actually relaxed the budget constraint and reduced the aggregate ‘‘life cycle deficit’’, the difference between aggregate labor income and aggregate consumption. At the same time, however, the age profile of consumption rose faster than the age profile of labor income over these years, and these changes more than offset the favorable demographic trend, leading to a large net increase in the life cycle deficit. In 2000 the life cycle deficit was paid more than fully through expenditure of private asset income, while the government was a net saver. In 2008, however, while private asset income was still used to fund the deficit, now the government dissaved, with a deficit of 4.5% of GDP. Looking to the future, the economic support ratio is projected to decline rapidly. Demographic forces will raise consumption by 2.25% per year and labor income by only 1.63% per year, an unsustainable situation. The unfortunate conclusion is that Spain is indeed living beyond its means, even in this recent period of favorable demography, and its real problems will begin in the coming decades as population aging occurs and the demographic situation deteriorates. In ‘‘Who supports the elderly? Changing economic lifecycle reallocation in Taiwan, 1985 and 2005’’, Lai and Tung consider the changes over a twenty year period in the old-age support system and the economic circumstances of the elderly. In 1985, the role of the public sector was very limited and, hence, the elderly had to rely on familial support, continued work, or assets. Family support was most important with inter- and intra-household transfers funding about half of the needs of the elderly. Continued employment funded about one-third of consumption by the elderly and one-fifth was funded by relying on assets. By one measure, at least, the support system appears to have been quite successful. Per capita consumption by the elderly was about 10% less than consumption by prime-age adults. The near equality of per capita consumption across adult ages is noteworthy for an economy that experienced such rapid economic growth in the
preceding decades. The labor income of young and middle-aged adults was far higher than that available to the elderly during their work lives. The decades after 1985 were marked by important social and economic changes in Taiwan. Strong economic growth continued, co-residence of adult children and their parents declined, and important public programs were introduced. The Old-Age Allowance program introduced in 1993 and National Health Insurance in 1995 led to a substantial increase in the importance of net public transfers to the elderly. Total and publicly-funded spending on health increased substantially among the elderly, but otherwise standards of living among the elderly do not appear to have increased relative to younger generations. The old-age support system did change in other important ways. The elderly of 2005 relied much less on income from their own labor and familial transfers and much more on asset-based reallocations. The decline in familial transfers was due entirely to declining intra-household transfers among co-residing adults with no compensating rise in inter-household transfers. The elderly in 2005 had much higher asset income than the elderly of 1985. That income was more than sufficient to fund a larger share of consumption and, consequently, the elderly in 2005 continued to save and at a much higher rate relative to consumption than the elderly in 1985. Contrary to the standard lifecycle model, the elderly did not support old-age consumption by dis-saving. Whether expanding transfer systems crowded out saving is unclear. Saving rates declined at working ages between 1985 and 2005 and increased at older ages. Perhaps lower saving among workers was a consequence of expanded support systems, but other factors could be responsible. Racelis, Abrigo, and Salas, in ‘‘Financing Consumption of the Lifecycle and Overseas Workers’ Remittances’’, make two important contributions to understanding the generational economy in a developing country context. One important part of their analysis is their exploration of how the generational economy has changed between 1999 and 2007. The Philippines experienced solid economy growth during this period, but unlike Taiwan the role of the public sector did not change providing little support for the elderly. In the absence of public support, the elderly in the Philippines relied entirely on their own labor and asset-based reallocations. At first glance, it seems puzzling that the net private transfers to the elderly were so small in contrast to their important role in Taiwan. The explanation is that the pre-transfer resource gap between the elderly and their adult children was much smaller in the Philippines than in Taiwan. The private transfer inflows and outflows were actually larger in the Philippines, relative to consumption, than in Taiwan, but in the Philippines the downward and upward flows were balanced while in Taiwan they heavily favored the elderly. The second issue addressed by Racelis, Abrigo, and Salas is the treatment of remittances in national accounts and in NTA. This is very important in the Philippines because so many Filipinos are working overseas. In the Philippines, many of these employees are contract workers and classified as temporarily abroad. They are classified as residents of the Philippines and their earnings are included in GDP following the standard procedures of the UN and the Philippine System of National Accounts. This approach is preferred for NTA and the earnings of these workers should be included in labor income. An alternative approach is to classify this income as transfers from the rest of the world. The line between what is classified as labor income and what is classified as transfers from the rest of the world is inevitably arbitrary, however. The classification percolates through NTA in ways described in useful detail by the authors. The issue they raise is more than a definitional issue because the measured impact of international labor migration depends on the particular nature of labor contracts or the lack thereof. A much smaller portion of the labor income
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
earned by Mexican migrants residing in the US, for example, is classified as labor income and a greater portion as transfers from ROW, because Mexican immigrants are much less likely to be contract workers. Finally, the paper by Albis et al. ‘‘The lifecycle deficit in France 1979–2005’’ presents a time series for the consumption and labor income profiles for France, which is one of the most recent European NTA countries. This is the main novelty of this paper. Being one of the main European countries these estimates give interesting insights on how the lifecycle deficit evolves along time. The authors first obtain the deficit profile for 2005 and compare it with other countries. Second, they obtain a time series of the components of the life cycle deficit – consumption both private and public and labor income. The 2005 estimates show a quite stable pattern for the cross sectional consumption profile. It stays almost constant once adulthood is reached, especially after retirement. In this respect, France differs from the United States and many other developed countries, in which per capita consumption increases after age 60 or 65. This is due to the increase in health consumption either public or private, as in Sweden or the US. Public health consumption is also growing in France in old age, but the decreasing pattern of private consumption compensates for this, resulting in a flat pattern for total consumption. Still, part of the difference can be explained by the fact that a substantial part of long-term-care expenditure is funded by cash transfers, which are therefore not counted as public consumption. Comparison of the labor income profile shows that France is characterized by a small proportion of total lifecycle income earned by persons under age 20 or ages 65 and over. This is explained by a relatively long duration of education on average and by a relatively low employment rate at older ages. The resulting life cycle deficit covers (25 years during youth and 22 years during old age for a total of 47 years, while labor income exceeds consumption during 32 years (from age 26 to 58). The evolution of consumption and labor income age profiles beginning from 1979 is then analyzed. The age at which labor income peaks shifts by 10 years during the period, from 39 years in 1979 to 49 in 2005. Most of shift occurs during the first decade. This trend seems to be mainly driven by a significant reduction in the share of employed senior workers, particularly due to the various measures implemented for earlier retirement in the early 1980s. The age profile of consumption is shifting upward relative to labor income during this period. The shape of the profile is relatively stable, apart from a slight shift upward for people aged 60 and over compared to the younger population. The resulting trend of the life cycle deficit implies a shortening of the period during which an individual is a net contributor. Another trend affecting the evolution of the life cycle deficit is the improvement in survival, especially at older ages. As a result of both the reduction of the productive period and the increases in life expectancy, the aggregate lifecycle deficit increases. This value expressed as a share of GDP of the same year, goes from a 0.47% of GDP in 1979, to a 10.23% of GDP in 2005. Comparing this ratio considering only the lifecycle deficit of the old (over 60) or the young (under 20) shows that the latter is increasing relatively to the former. Bernhard Hammer, Alexia Fuernkranz-Prskawetz, and Inga Freund contribute an innovative comparative analysis in their paper ‘‘Production Activities and Economic Dependency by Age and Gender in Europe: A Cross-Country Comparison’’ based on NTA and NTTA estimates by gender for Europe. The authors offer a new measure, the economic dependency ratio, defined as the aggregate life cycle deficit (LCD) as a share of aggregate labor income. The measure is a refinement of the purely demographic dependency ratio in that it takes into account not only the population age structure, but also the real production and consumption levels at each age. Quite interestingly, it allows for flexible age
5
limits, instead of fixing a priori the age limits of working (15 and 64, for example). In a further extension, the paper extends the measure to incorporate gender differences in both the market based NTA profiles and the non-market based NTTA profiles. One of the contributions of the paper is its assessment of using a cross country comparable data set, the EU-SILC, to construct NTA estimates. The paper shows how NTA framework provides a clear measure of the economic impact of not only population age structure but also the different life cycle patterns. In particular, they show how early entry in the labor market and relatively low consumption of children improves the LCD for Germany and Austria. The same goal is reached in Sweden through a late exit from the labor market, while both Sweden and Slovenia have a high contribution of women in paid work. The introduction of unpaid work (NTTA) also leads to interesting conclusions. First, the old contribute substantially to unpaid work, but they are also high consumers. Second, one can compute the monetary value of the pension many European females are experiencing. Higher participation in labor market improves the monetary LCD, but at the cost of shrinking unpaid production. Besides the obvious conclusions on the potential reduction of the gender gap, which in general does not disappear completely, this demonstrates the importance of taking monetary and non-monetary transfers into account in policy design. The contribution of women to total labor income is among the highest within Europe in Sweden and Slovenia while Italy and the UK are at the other extreme. Finally, Gal, Szabo, and Vargha contest conventional views about intergenerational transfers in their paper, ‘‘The age-profile of invisible transfers: The true size of asymmetry in inter-age reallocations.’’ Transfers are very poorly measured in the system of national accounts and other economic statistics. As a consequence, they are poorly suited for understanding the magnitude and nature of economic flows to children and the elderly. Two gaps are particularly serious. The first is that only private transfers between households are captured, while transfers within households are not. NTA estimates consistently show that transfers within household dwarf transfers between households. The second gap is the exclusion of non-market time that fails to value time spent mostly by women in childrearing and caregiving for elderly relatives. Nonmarket time supplied by the elderly, caring for grandchildren for example, may also be important. This problem is addressed using National Time Transfer Accounts (NTTA). Using NTA and NTTA estimates for Hungary, the authors demonstrate the fundamental asymmetry in the institutional approaches to supporting children and the elderly. They express it as follows: ‘‘To put it sharply, children are raised by their parents, the elderly rely on society.’’ They find in Hungary ‘‘that two-thirds of the full transfer package for children is exchanged within the household’’ and, hence, goes uncaptured by public statistics. In contrast, ‘‘90 percent of the inter-age transfers for the elderly go through public channels.’’ Understanding this asymmetry is essential when interpreting the observation that in rich countries public transfers to the elderly are much greater than public transfers to children.2 As the authors point out, this merely documents that transfers to the elderly are socialized and to children are privatized. Moreover, they find that the costs of children are higher than the cost of the elderly once the ‘‘invisible transfers’’ are revealed. The papers presented in this Special Issue show the scope for fruitful interdisciplinary research collaboration between demography and economics. In fact intergenerational transfers are in the 2 Abio et al. (2015) observe that this puzzle is present in most countries in which the welfare state is sizable; despite the fact that market can more easily substitute transfers to the old that transfers to the young.
6
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
core of the interplay between demography and economy. The papers in this issue illustrate this point for different countries and raise a number of issues that provide interesting insights for these interactions. Given the dynamic nature of this issue, it would be valuable to have longitudinal data on the role of transfers in the economy, but the cross sectional estimates employed here and the various attempts to carry out longitudinal analysis are already quite informative. For example, the paper by Patxot et al. investigates the sustainability of the recent patterns of consumption in Spain in the face of the rapid population aging that will soon occur there, a topic that depends on both public and private reallocation systems. The retrospective longitudinal analysis done for Taiwan shows how family transfers are gradually substituted by the market and the government. This phenomenon occurs as the economy develops, but at different speeds, as seen for the case of the Philippines. Children must rely on transfers to fund their consumption because they cannot make use of asset markets. By contrast, the elderly can very readily make use of asset markets to fund their old age consumption through life cycle saving. It is surprising, then, that government transfers tend to focus on financing the needs of the elderly much more than those of children. As pointed out by Abio et al. (2015 (forthcoming)), this puzzle is present in most countries. The application to the Hungarian case shows that the introduction of time transfers worsens this picture. This should not necessarily be understood as a conflict among generations but as a way to balance the transfers received and benefits given along the lifecycle. In other words, it is as a way to avoid intra generational differences in income distribution linked to the reality of how children are raised. The paper by Hammer et al. concludes that welfare state reform needs to take into account both time and monetary transfers. These, together with the need to consider the balance of government intervention on forward and backward intergenerational transfers seems a good roadmap to facing challenges of keeping the welfare state in an ageing world and global world. In fact lessons derived from the design of the different welfare state models
in the developed world seem worth considering when designing the social protection system in the developing world. The papers in this journal illustrate the application of NTA to a wide range of questions and problems, but there is still much more to be done. NTA and NTTA provide a new and more detailed map of hitherto largely unexplored economic territory, and the process of understanding the implications of what is revealed is ongoing. References Abio, G., Patxot, C., Renteria, E., Souto, G., 2015. Towards a Balanced Welfare State, In: Gas-Aixendri, M., Cavallotti, R. (Eds.), Family and Sustainable Development. Thomson Reuters Aranzadi, (forthcoming). Lee, R., Donehower, G., 2011. Private transfers in comparative perspective. Population Aging and the Generational Economy: A Global Perspective. In: Lee, R., Mason, A. (Eds.). Edward Elgar, Cheltenham, UK and Northampton, MA, pp. 185–208. Lee, R., Mason, A., et al, 2014. Is low fertility really a problem? Population aging, dependency, and consumption. Science 346 (6206), 229–234. Lee, R., A. Mason, principal authors and editors (2011). Population Aging and the Generational Economy: A Global Perspective. Edward Elgar, Cheltenham, UK. Patxot, C., Rentería, E., et al, 2012. Measuring the balance of government intervention on forward and backward family transfers using NTA estimates: the modified Lee Arrows. Int. Tax Public Finance 19 (3), 442–461. United Nations Population Division, 2013. National Transfer Accounts Manual: Measuring and Analysing the Generational Economy. United Nations, New York.
Concepció Patxot Department of Economic, University of Barcelona, Spain Ronald Lee Departments of Demography and Economic Theory, University of California, Berkeley, United States Andrew Mason Department of Economics, University of Hawaii and East-West Center, United States Available online 4 May 2015
Contents lists available at ScienceDirect
The Journal of the Economics of Ageing journal homepage: www.elsevier.com/locate/jeoa
Introduction: Special issue on exploring the generational economy
The interplay between demography and economics is a rich and important area of interdisciplinary study involving researchers from economics, demography, sociology, political science, anthropology, psychology, and other fields of study. In the past, economic analysis incorporated demography using highly stylized models. Many studies and even the System of National Accounts takes the household as the basic decision-making unit treating the household as a black box. This obscures many important aspects of economic behavior, such as, the roles and economic effects of children, the elderly, and gender. In the past, macroeconomic models ignored population age structure or relied on highly simplified models of age structure. Population and labor force were treated as equivalent or populations were assumed to consist of only two or three adult age groups. These highly simplified representations of the world have been very valuable and yielded important insights about intergenerational transfers, credit markets, economic growth, and other macroeconomic processes, but they have obvious limitations. Research on the interplay between demography and economics has shifted over time to using more realistic models of the generational economy. In part, this is a natural outgrowth of the increase in computer power available on researchers’ desks (or in their laps). In part, it reflects the need to provide richer accounts of the demographic transition being experienced in almost every developing and developed country. Since the 1970s Family Economics has extensively explored decision-making within the family about childbearing, marriage, and divorce. The role of altruism and other motives for intergenerational transfers have been extensively explored. Models have also bridged the gap between the micro and macro levels of analysis, leading to a body of work sometimes called ‘‘family macro’’. The demographic transition has led to dramatic improvement in life expectancy that is an outcome, in part, of economic change but also represents an increase in welfare that is poorly captured by many economic models and economic statistics. The demographic transition has led to sharp changes in population age structure with more to come that are having a pervasive influence on our economies. The papers in this special issue were selected from among those presented at the Ninth Global Meeting of the Working Group on Macroeconomic Aspects of Intergenerational Transfers hosted by the University of Barcelona and held at Faculty of Economics and Business, 3–8 June 2013. The conference was attended by members of the National Transfer Account network and other researchers working to improve our understanding of the connections between economy and demography.
http://dx.doi.org/10.1016/j.jeoa.2015.04.001 2212-828X/Ó 2015 Elsevier B.V. All rights reserved.
An overview of National Transfer Accounts National Transfer Accounts (NTA) provides a conceptual framework and a system of accounts for the generational economy (Lee and Mason, 2011). The generational economy refers to the social institutions and economic mechanisms used by generations or age groups to produce, consume, share and save resources. Understanding the generational economy is important due to a fundamental feature of all societies – the economic lifecycle. Over an extended period at the beginning of life in all societies, children consume more than they produce through their labor – a gap we refer to as a life cycle deficit. This is followed by a lifecycle surplus over many years during which humans produce more through their labor than they consume. Over the final phase of the economic lifecycle in all contemporary societies for which we have estimates (other than some hunter–gatherer and subsistence swidden groups), humans again experience a lifecycle deficit consuming more than they produce through their labor. The significance of the economic lifecycle depends on the economic mechanisms and institutions that enable its existence. A very substantial portion of every nation’s resources must be reallocated across age – from the lifecycle surplus ages to the lifecycle deficit ages. The needs of children and youth are met through transfers – private transfers from their families and public transfers through government with the relative mix varying across countries. To a limited extent, young adults may rely on credit markets (student loans, consumer debt). The picture is very different for older adults. The relative importance of public and private transfers to the elderly varies greatly from country to country. Moreover, in many countries the elderly rely heavily on assets (asset income or dis-saving) to fund their lifecycle deficits. NTA has been developed to complement National Accounts by adding an age dimension, and in most respects is consistent with National Accounts. The focus on the individual in NTA rather than the household, however, leads to a much richer description of public and private transfers than in National Accounts. NTA provides estimates of transfers by purpose and by the age of the provider and the recipient. Estimates of both inter- and intra-household transfers are constructed. Moreover, NTA provides a consistent set of data on how asset income, debt, and assets vary over the lifecycle. NTA are constructed using a variety of sources of data from National Accounts, administrative records that document taxes and public spending, household surveys that contain information about income, consumption, and demographic characteristics, labor force surveys, and population censuses. More recently, time
2
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
use surveys are used in many countries to construct estimates of home production activities including child and elder care time. The methods for constructing basic accounts are described in the NTA Manual recently published by the United Nations Population Division United Nations Population Division (2013). Research teams in over 50 countries have constructed or are currently constructing NTA, sometimes for multiple years. Estimates for at least some components of basic NTA are available for 40 countries on the NTA website (Lee and Mason, 2014). The initial goal of the NTA network was to construct and analyze accounts for a recent year in countries that varied greatly in their political and economic systems, level of development, quality of data, and so forth. This collaborative effort allowed us to evaluate and to refine NTA methods. It also proved to be very useful for improving our understanding of the generational economy and the macroeconomic implications of changing population age structure. The contribution by Oosthuizen on the demographic dividend in South Africa is an excellent example of this kind of analysis. Many of the papers in this volume employ NTA data for multiple years to explore how features of the generational economy are evolving over time as populations age and as economies develop or respond to economic crisis. This is an important development and will lead to many research opportunities for exploring synthetic cohort data. Vogt and Kluge use NTA time series data of this sort to look at the role of West–East transfers in the convergence in life expectancy after reunification. The Patxot et al. paper decomposes changes in NTA across time, with the purpose of testing the sustainability of the per capita levels of Spanish consumption before the 2007 international financial crisis in face of population ageing. The paper by Racelis, Abrigo, and Salas compares the NTA estimates for 1999 and 2007 for the Philippines exploring also the role of migrant’s remittances, quite relevant in this country. The paper by Lai and Tung takes a longer perspective by investigating changes in the structure of transfers from 1985 to 2005. Finally, Albis et al. take a longer term perspective to analyze the evolution of the French lifecycle deficit for more than two decades (1979– 2005). The value of NTA can be enhanced through other extensions that are explored here. One active areas of research is to explore how the lifecycle and the age reallocation system varies by socioeconomic groups. NTA values in the basic accounts are aggregates or per capita values (averages). This is ideal for some applications, but for other applications it is important to know how lifecycle issues, intergenerational transfers, systems of support, etc. influence low- or middle-income groups. Papers by Mejia-Guevara and by Jimenez-Fontana introduce disaggregation by socioeconomic status (mainly measured by education) for Mexico and Costa Rica, respectively. Another important area of work is to use a broader measure of economic flows incorporating the value of goods and services that are produced primarily at home. Time transfers are a very important part of the intergenerational support system. Mothers, in particular, provide valuable services through childrearing, caregiving, and other activities that occur outside the market place. Estimating National Time Transfer Accounts (NTTA) provides a more realistic view of the costs of children, the economic contributions of women, and the costs and contributions of the elderly. The development of NTTA has been essential to another NTA-related enterprise – estimating accounts separately by gender. The papers by Zannella, Jimenez-Fontana, and Gal et al. illustrate this approach for the cases of Italy, Costa Rica and Hungary, respectively. Many of the papers in this special issue explore the implications of incorporating a more complete assessment of the economic value of work and the intergenerational transfers that arise primarily between family members. Hammer et al. provides a comparative analysis on the economic dependency ratio in eight
European countries using both the NTA profiles and a first approximation of the NTTA profiles. Going deeply into the role of transfers, the paper by Gal, Szbo, and Vargha uses NTA and NTTA profiles from Hungary to show the extent to which the cost of the elderly and of the children are socialized. The authors extend the work previously done in other countries by Lee and Donehower (2011) and Patxot and Rentería (2012) using NTA profiles, to include the time transfers (NTTA) in the picture.1 This reinforces the result that public transfers to children are far less important than public transfers to the old. This phenomenon, which occurs in other countries, is quite surprising if one takes into account that transfers to the old given to finance long-term care, pensions and health care, are by far more substitutable in the market than transfers to children. The first two papers in the special issue, authored by Tobias Vogt and Fanny Kluge and by Morne Oosthuizen, were selected as the outstanding papers presented at the conference, a prize sponsored by the Journal of the Economics of Aging and Elsevier. The paper by Vogt and Kluge ‘‘Can public spending reduce mortality disparities? Findings from East Germany after Reunification’’ investigates the impact of public spending on mortality disparities. The paper mainly focuses on the German case. The reunification of East and West Germany in 1990 and the improved conditions in the East thereafter provide a natural experiment that is used in this paper to evaluate the effect on mortality at older ages of increased pension benefits and increased health care spending. Before reunification, the gap between life expectancy in the East and West was large and growing. After reunification, life expectancy at 65 in the East increased rapidly and converged with that in the West, in contrast to Czechoslovakia, Hungary and Poland where mortality conditions are still similar to those in East Germany in 1990. This suggests that greatly increased financial transfers and spending on health care following reunification with West Germany played an important role. To investigate this idea more formally, a difference-in-differences analysis is used, finding that a one euro expenditure on health or pensions for older people in the East raised the length of life by 3 h. In further analysis, equations are estimated separately in East and West for the log of age and gender specific mortality by cause, where the covariates include average health care and pension expenditures by age, sex, region and period. The results show that expenditures on both pensions and health care reduce mortality, but health care is most important. Furthermore, expenditures in the East led to greater improvements than similar expenditures in the West. Morne Oosthuizen, in his paper ‘‘Bonus or Mirage? South Africa’s Demographic Dividend’’, explores a very important issue throughout the developing world – how best to achieve more rapid economic growth by harnessing the demographic dividend. The South African experience is important in many respects. Its fertility has declined further than elsewhere in the region and South Africa has had to deal with many serious problems arising as a consequence of apartheid. Oosthuizen addresses the first demographic dividend showing how the support ratio in South Africa is influenced by the particular features of labor income and consumption. Realizing the first dividend in South Africa has been a challenge because labor income is so low among young adults, a feature that has been found in some other African countries as well. Consumption patterns have also had an important influence on the support ratio. As Oosthuizen observes, however, it is unclear what policies might be pursued to influence consumption patterns in contrast to how low youth productivity might be addressed. Oosthuizen goes on to explore the second demographic dividend – how demographic change can be translated into higher
1
Patxot et al. (2012) discuss the related economic literature.
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
capital accumulation and investment in human capital. Spending on human capital in South Africa appears to reflect the quantity– quality tradeoff. Average public and private health and education spending per child is very similar to the levels found in other countries with similar rates of childbearing and similar levels of development. An important issue addressed by Oosthuizen is that human capital spending in South Africa is ineffective because of its failure to address the great disparities that persist. If this problem is not addressed, South Africa may fail to capture one of the most powerful pro-development channels of the second demographic dividend. The next three papers employ extensions of NTA data to document and to analyze important features of the generational economy in different contexts. Mejia Guevara, in his paper ‘‘Economic Inequality and Intergenerational Transfers: Evidence from Mexico’’, extends the standard NTA analysis by introducing socioeconomic status and by comparing NTA profiles at two points in time 1994 and 2004. The main issue addressed by this paper is the substantial decline in inequality over this period. The expansion of transfers, in general, and the conditional cash transfer program Progresa– Oportunidades, contributed in many ways including the expansion of basic education. The paper is interesting from a methodological perspective as the author disaggregates NTA estimates by socioeconomic status, using educational attainment as a proxy, in order to investigate if socioeconomic status influences age reallocations. As in the paper by Jimenez-Fontana, the decomposition of NTA profiles by education status, allows us to see differential patterns in the country, some of which resemble those observed in developed countries. For example, the most educated among the elderly show consumption increasing with age, partly due to the increase in public consumption, as in some developed countries, while labor income is not yet reduced by early retirement. The insights of the author turn out to be very instructive when looking at the age reallocations at different education levels. In brief, the transfer system seems to be acting as redistributive tool that operates most completely to the benefit of the less educated. ‘‘On the one hand, while for the less educated the main financing source of child consumption is public transfers (48.0), followed by private transfers (39.8%) and labor income (11.9%); private transfers are the main financing source for the most educated (96.7%). On the other hand, regarding consumption of the elderly in the lowest SES group, consumption is funded by asset-based reallocations (36.4%), public transfers (37.9%) and labor income (33.2%), while it is funded entirely by asset-based reallocations (108.2%) in the highest socioeconomic group’’. Despite the development of asset markets, intrafamily arrangements still play a role in Mexico. Quite interestingly, the elderly in the upper SES group rely on their assets, and transfer a substantial amount of resources to their children and grandchildren in the form of private transfers. This crowding out effect is observed in all socioeconomic groups, but the magnitude is different, ranging from 7.5% in the less educated to 44.2% for the most educated. The magnitudes have also increased from 1994 to 2004 as a result of the increase in transfers during the period. The paper ‘‘Reallocation of Resources between Generation and Genders in the Market and Non-Market Economy. The Case of Italy’’ by Marina Zannella provides a complete picture of intergenerational transfers in Italy combining both monetary transfers, both market and non-market transfers, with time transfers. The author begins by pointing out that investing in human capital of children and caring for the elderly are the two major tasks for intergenerational transfers. In Northern Europe the public sector plays a major or dominant role in both whereas in Southern Europe the family plays a greater role, particularly for children.
3
This paper examines the system of intergenerational reallocation of resources in Italy in 2008, including both market goods as in National Transfer Accounts, NTA, and also transfers of time as in the new National Time Transfer Accounts, NTTA. This permits a thorough examination of gender differences in production and consumption, since in Italy men are more active in market work while women, who have relatively low labor force participation, are more active in home production and care. This paper presents many interesting results. Starting with standard NTA by gender, the authors show that males and females consume very similar amounts at each age, except that after age 70 men begin to consume considerably more than women. Men have far higher labor income than women, roughly twice as great at each age. Consequently, while male and female children benefit equally from family transfers until age 18 or so, after this the net family transfers received by women are much greater than those by men, and indeed after age 25 men make substantial net familial transfers at every age whereas women continue to be net recipients at every age except in their forties when they make small net transfers to others. This is a misleading picture of women’s economic activity, however, since so much of female work is non-market activity in the home. The NTTA based on time use data shows that women produce roughly twice as much in home production after age 20, and girls begin producing more than boys even before they are ten years old. Transfers of home production are the mirror image of transfers of market goods, with females making large net transfers to males after age 20 while males are net recipients throughout their whole lives. In NTTA, children are much more costly and have a much larger life cycle deficit, than in NTA, due to their time costs. At other ages, the life cycle deficit for sexes combined appears very similar with and without home production included (NTTA vs NTA) because the ages of women providing home production and the men consuming it are similar, so it tends to average out. From birth through the mid-twenties, but particularly before age 5, familial transfers are much larger than public ones. Zanella suggests that Italy should provide greater public support for children and young adults, reducing the familial role in this regard. This would aid gender balance since the familial role is in fact largely female. Pamela Jimenez-Fontana, in ‘‘Analysis of Non-remunerated Production in Costa Rica’’, focuses on non-market production in Costa Rica which she estimates at 40% of GDP. The author presents NTA and NTTA profiles for males and females, finding heavy specialization by gender. In particular she finds that there is strong specialization by type of production, so that men specialize in the labor market and women specialize in home production. Hence, the analysis by gender implies that taking into account home production reduces the gender gap by half in Costa Rica. She also finds that there is specialization in the different tasks in home production. Women specialize in the production of care and other activities like cleaning, cooking, and laundry, while men specialize in car and farm maintenance and spend some time to travel to and from work. Overall, the activities that men specialize in represent a small proportion of all non-remunerated production. This situation is not surprising in a developing country like Costa Rica. The author is interested in the future evolution of labor division within the household, including help received from the rest of the family network. Hence, the author analyses to what extent gender specialization is motivated by a higher opportunity cost related to educational attainment. To that end, the NTTA profiles are disaggregated by education level. The results show that education matters: Women with high educational level dedicate less time to activities like cooking and cleaning, which may be explained by the high opportunity cost of staying at home doing these duties without economic compensation. But still, they spend
4
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
more time on these tasks than men. Regarding the production of care, there is also heterogeneity for both men’s and women’s profiles by educational level. For women, one can observe a first peak in the age profile of the production of care, which could be identified as taking care of children. Women with low levels of education have two initial peaks and women with completed high school or more only have one. These peaks are coherent with the observed TFR in Costa Rica by educational level. The production of care performed by women suggests a postponement of fertility’ and male’s profiles mainly follow this trend; women with high levels of education have the peaks of production of care later in life. Quite interestingly, the reduction of time spent in cooking and cleaning activities by females is compensated by an increase in time engaged in household management. Educated females, a growing group in Costa Rica, spend more time on these kinds of tasks, than women with primary education or less. In some sense, this trend represents women’s empowerment in the household and in the economy. Less educated women depend economically on their partners and family members. The next four papers focus on longitudinal issues and provide examples of how a future longitudinal NTA database can contribute to shed light on the role of intergenerational transfers. The first of these papers by Patxot, Renteria and Souto explores an important question in Spain and is described well by the title of their paper: ‘‘Can we keep the pre-crisis living standards? An analysis based on NTA profiles’’. Between 2000 and 2008, the period of analysis, demographic conditions were actually exceptionally favorable, due to the high proportion in the working ages and the low numbers of children and elderly. Using NTA data, the economic support ratio is shown to be at its peak around 2008, a situation which will be followed by decades of decline as the population ages. Furthermore, simulations show that demographic change from 2000 to 2008 actually relaxed the budget constraint and reduced the aggregate ‘‘life cycle deficit’’, the difference between aggregate labor income and aggregate consumption. At the same time, however, the age profile of consumption rose faster than the age profile of labor income over these years, and these changes more than offset the favorable demographic trend, leading to a large net increase in the life cycle deficit. In 2000 the life cycle deficit was paid more than fully through expenditure of private asset income, while the government was a net saver. In 2008, however, while private asset income was still used to fund the deficit, now the government dissaved, with a deficit of 4.5% of GDP. Looking to the future, the economic support ratio is projected to decline rapidly. Demographic forces will raise consumption by 2.25% per year and labor income by only 1.63% per year, an unsustainable situation. The unfortunate conclusion is that Spain is indeed living beyond its means, even in this recent period of favorable demography, and its real problems will begin in the coming decades as population aging occurs and the demographic situation deteriorates. In ‘‘Who supports the elderly? Changing economic lifecycle reallocation in Taiwan, 1985 and 2005’’, Lai and Tung consider the changes over a twenty year period in the old-age support system and the economic circumstances of the elderly. In 1985, the role of the public sector was very limited and, hence, the elderly had to rely on familial support, continued work, or assets. Family support was most important with inter- and intra-household transfers funding about half of the needs of the elderly. Continued employment funded about one-third of consumption by the elderly and one-fifth was funded by relying on assets. By one measure, at least, the support system appears to have been quite successful. Per capita consumption by the elderly was about 10% less than consumption by prime-age adults. The near equality of per capita consumption across adult ages is noteworthy for an economy that experienced such rapid economic growth in the
preceding decades. The labor income of young and middle-aged adults was far higher than that available to the elderly during their work lives. The decades after 1985 were marked by important social and economic changes in Taiwan. Strong economic growth continued, co-residence of adult children and their parents declined, and important public programs were introduced. The Old-Age Allowance program introduced in 1993 and National Health Insurance in 1995 led to a substantial increase in the importance of net public transfers to the elderly. Total and publicly-funded spending on health increased substantially among the elderly, but otherwise standards of living among the elderly do not appear to have increased relative to younger generations. The old-age support system did change in other important ways. The elderly of 2005 relied much less on income from their own labor and familial transfers and much more on asset-based reallocations. The decline in familial transfers was due entirely to declining intra-household transfers among co-residing adults with no compensating rise in inter-household transfers. The elderly in 2005 had much higher asset income than the elderly of 1985. That income was more than sufficient to fund a larger share of consumption and, consequently, the elderly in 2005 continued to save and at a much higher rate relative to consumption than the elderly in 1985. Contrary to the standard lifecycle model, the elderly did not support old-age consumption by dis-saving. Whether expanding transfer systems crowded out saving is unclear. Saving rates declined at working ages between 1985 and 2005 and increased at older ages. Perhaps lower saving among workers was a consequence of expanded support systems, but other factors could be responsible. Racelis, Abrigo, and Salas, in ‘‘Financing Consumption of the Lifecycle and Overseas Workers’ Remittances’’, make two important contributions to understanding the generational economy in a developing country context. One important part of their analysis is their exploration of how the generational economy has changed between 1999 and 2007. The Philippines experienced solid economy growth during this period, but unlike Taiwan the role of the public sector did not change providing little support for the elderly. In the absence of public support, the elderly in the Philippines relied entirely on their own labor and asset-based reallocations. At first glance, it seems puzzling that the net private transfers to the elderly were so small in contrast to their important role in Taiwan. The explanation is that the pre-transfer resource gap between the elderly and their adult children was much smaller in the Philippines than in Taiwan. The private transfer inflows and outflows were actually larger in the Philippines, relative to consumption, than in Taiwan, but in the Philippines the downward and upward flows were balanced while in Taiwan they heavily favored the elderly. The second issue addressed by Racelis, Abrigo, and Salas is the treatment of remittances in national accounts and in NTA. This is very important in the Philippines because so many Filipinos are working overseas. In the Philippines, many of these employees are contract workers and classified as temporarily abroad. They are classified as residents of the Philippines and their earnings are included in GDP following the standard procedures of the UN and the Philippine System of National Accounts. This approach is preferred for NTA and the earnings of these workers should be included in labor income. An alternative approach is to classify this income as transfers from the rest of the world. The line between what is classified as labor income and what is classified as transfers from the rest of the world is inevitably arbitrary, however. The classification percolates through NTA in ways described in useful detail by the authors. The issue they raise is more than a definitional issue because the measured impact of international labor migration depends on the particular nature of labor contracts or the lack thereof. A much smaller portion of the labor income
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
earned by Mexican migrants residing in the US, for example, is classified as labor income and a greater portion as transfers from ROW, because Mexican immigrants are much less likely to be contract workers. Finally, the paper by Albis et al. ‘‘The lifecycle deficit in France 1979–2005’’ presents a time series for the consumption and labor income profiles for France, which is one of the most recent European NTA countries. This is the main novelty of this paper. Being one of the main European countries these estimates give interesting insights on how the lifecycle deficit evolves along time. The authors first obtain the deficit profile for 2005 and compare it with other countries. Second, they obtain a time series of the components of the life cycle deficit – consumption both private and public and labor income. The 2005 estimates show a quite stable pattern for the cross sectional consumption profile. It stays almost constant once adulthood is reached, especially after retirement. In this respect, France differs from the United States and many other developed countries, in which per capita consumption increases after age 60 or 65. This is due to the increase in health consumption either public or private, as in Sweden or the US. Public health consumption is also growing in France in old age, but the decreasing pattern of private consumption compensates for this, resulting in a flat pattern for total consumption. Still, part of the difference can be explained by the fact that a substantial part of long-term-care expenditure is funded by cash transfers, which are therefore not counted as public consumption. Comparison of the labor income profile shows that France is characterized by a small proportion of total lifecycle income earned by persons under age 20 or ages 65 and over. This is explained by a relatively long duration of education on average and by a relatively low employment rate at older ages. The resulting life cycle deficit covers (25 years during youth and 22 years during old age for a total of 47 years, while labor income exceeds consumption during 32 years (from age 26 to 58). The evolution of consumption and labor income age profiles beginning from 1979 is then analyzed. The age at which labor income peaks shifts by 10 years during the period, from 39 years in 1979 to 49 in 2005. Most of shift occurs during the first decade. This trend seems to be mainly driven by a significant reduction in the share of employed senior workers, particularly due to the various measures implemented for earlier retirement in the early 1980s. The age profile of consumption is shifting upward relative to labor income during this period. The shape of the profile is relatively stable, apart from a slight shift upward for people aged 60 and over compared to the younger population. The resulting trend of the life cycle deficit implies a shortening of the period during which an individual is a net contributor. Another trend affecting the evolution of the life cycle deficit is the improvement in survival, especially at older ages. As a result of both the reduction of the productive period and the increases in life expectancy, the aggregate lifecycle deficit increases. This value expressed as a share of GDP of the same year, goes from a 0.47% of GDP in 1979, to a 10.23% of GDP in 2005. Comparing this ratio considering only the lifecycle deficit of the old (over 60) or the young (under 20) shows that the latter is increasing relatively to the former. Bernhard Hammer, Alexia Fuernkranz-Prskawetz, and Inga Freund contribute an innovative comparative analysis in their paper ‘‘Production Activities and Economic Dependency by Age and Gender in Europe: A Cross-Country Comparison’’ based on NTA and NTTA estimates by gender for Europe. The authors offer a new measure, the economic dependency ratio, defined as the aggregate life cycle deficit (LCD) as a share of aggregate labor income. The measure is a refinement of the purely demographic dependency ratio in that it takes into account not only the population age structure, but also the real production and consumption levels at each age. Quite interestingly, it allows for flexible age
5
limits, instead of fixing a priori the age limits of working (15 and 64, for example). In a further extension, the paper extends the measure to incorporate gender differences in both the market based NTA profiles and the non-market based NTTA profiles. One of the contributions of the paper is its assessment of using a cross country comparable data set, the EU-SILC, to construct NTA estimates. The paper shows how NTA framework provides a clear measure of the economic impact of not only population age structure but also the different life cycle patterns. In particular, they show how early entry in the labor market and relatively low consumption of children improves the LCD for Germany and Austria. The same goal is reached in Sweden through a late exit from the labor market, while both Sweden and Slovenia have a high contribution of women in paid work. The introduction of unpaid work (NTTA) also leads to interesting conclusions. First, the old contribute substantially to unpaid work, but they are also high consumers. Second, one can compute the monetary value of the pension many European females are experiencing. Higher participation in labor market improves the monetary LCD, but at the cost of shrinking unpaid production. Besides the obvious conclusions on the potential reduction of the gender gap, which in general does not disappear completely, this demonstrates the importance of taking monetary and non-monetary transfers into account in policy design. The contribution of women to total labor income is among the highest within Europe in Sweden and Slovenia while Italy and the UK are at the other extreme. Finally, Gal, Szabo, and Vargha contest conventional views about intergenerational transfers in their paper, ‘‘The age-profile of invisible transfers: The true size of asymmetry in inter-age reallocations.’’ Transfers are very poorly measured in the system of national accounts and other economic statistics. As a consequence, they are poorly suited for understanding the magnitude and nature of economic flows to children and the elderly. Two gaps are particularly serious. The first is that only private transfers between households are captured, while transfers within households are not. NTA estimates consistently show that transfers within household dwarf transfers between households. The second gap is the exclusion of non-market time that fails to value time spent mostly by women in childrearing and caregiving for elderly relatives. Nonmarket time supplied by the elderly, caring for grandchildren for example, may also be important. This problem is addressed using National Time Transfer Accounts (NTTA). Using NTA and NTTA estimates for Hungary, the authors demonstrate the fundamental asymmetry in the institutional approaches to supporting children and the elderly. They express it as follows: ‘‘To put it sharply, children are raised by their parents, the elderly rely on society.’’ They find in Hungary ‘‘that two-thirds of the full transfer package for children is exchanged within the household’’ and, hence, goes uncaptured by public statistics. In contrast, ‘‘90 percent of the inter-age transfers for the elderly go through public channels.’’ Understanding this asymmetry is essential when interpreting the observation that in rich countries public transfers to the elderly are much greater than public transfers to children.2 As the authors point out, this merely documents that transfers to the elderly are socialized and to children are privatized. Moreover, they find that the costs of children are higher than the cost of the elderly once the ‘‘invisible transfers’’ are revealed. The papers presented in this Special Issue show the scope for fruitful interdisciplinary research collaboration between demography and economics. In fact intergenerational transfers are in the 2 Abio et al. (2015) observe that this puzzle is present in most countries in which the welfare state is sizable; despite the fact that market can more easily substitute transfers to the old that transfers to the young.
6
Editorial / The Journal of the Economics of Ageing 5 (2015) 1–6
core of the interplay between demography and economy. The papers in this issue illustrate this point for different countries and raise a number of issues that provide interesting insights for these interactions. Given the dynamic nature of this issue, it would be valuable to have longitudinal data on the role of transfers in the economy, but the cross sectional estimates employed here and the various attempts to carry out longitudinal analysis are already quite informative. For example, the paper by Patxot et al. investigates the sustainability of the recent patterns of consumption in Spain in the face of the rapid population aging that will soon occur there, a topic that depends on both public and private reallocation systems. The retrospective longitudinal analysis done for Taiwan shows how family transfers are gradually substituted by the market and the government. This phenomenon occurs as the economy develops, but at different speeds, as seen for the case of the Philippines. Children must rely on transfers to fund their consumption because they cannot make use of asset markets. By contrast, the elderly can very readily make use of asset markets to fund their old age consumption through life cycle saving. It is surprising, then, that government transfers tend to focus on financing the needs of the elderly much more than those of children. As pointed out by Abio et al. (2015 (forthcoming)), this puzzle is present in most countries. The application to the Hungarian case shows that the introduction of time transfers worsens this picture. This should not necessarily be understood as a conflict among generations but as a way to balance the transfers received and benefits given along the lifecycle. In other words, it is as a way to avoid intra generational differences in income distribution linked to the reality of how children are raised. The paper by Hammer et al. concludes that welfare state reform needs to take into account both time and monetary transfers. These, together with the need to consider the balance of government intervention on forward and backward intergenerational transfers seems a good roadmap to facing challenges of keeping the welfare state in an ageing world and global world. In fact lessons derived from the design of the different welfare state models
in the developed world seem worth considering when designing the social protection system in the developing world. The papers in this journal illustrate the application of NTA to a wide range of questions and problems, but there is still much more to be done. NTA and NTTA provide a new and more detailed map of hitherto largely unexplored economic territory, and the process of understanding the implications of what is revealed is ongoing. References Abio, G., Patxot, C., Renteria, E., Souto, G., 2015. Towards a Balanced Welfare State, In: Gas-Aixendri, M., Cavallotti, R. (Eds.), Family and Sustainable Development. Thomson Reuters Aranzadi, (forthcoming). Lee, R., Donehower, G., 2011. Private transfers in comparative perspective. Population Aging and the Generational Economy: A Global Perspective. In: Lee, R., Mason, A. (Eds.). Edward Elgar, Cheltenham, UK and Northampton, MA, pp. 185–208. Lee, R., Mason, A., et al, 2014. Is low fertility really a problem? Population aging, dependency, and consumption. Science 346 (6206), 229–234. Lee, R., A. Mason, principal authors and editors (2011). Population Aging and the Generational Economy: A Global Perspective. Edward Elgar, Cheltenham, UK. Patxot, C., Rentería, E., et al, 2012. Measuring the balance of government intervention on forward and backward family transfers using NTA estimates: the modified Lee Arrows. Int. Tax Public Finance 19 (3), 442–461. United Nations Population Division, 2013. National Transfer Accounts Manual: Measuring and Analysing the Generational Economy. United Nations, New York.
Concepció Patxot Department of Economic, University of Barcelona, Spain Ronald Lee Departments of Demography and Economic Theory, University of California, Berkeley, United States Andrew Mason Department of Economics, University of Hawaii and East-West Center, United States Available online 4 May 2015