Discussion – Week 8 Top of Form Discussion: Sociocultural Differences in Perspectives on Aging Some cultures view death not

AGING, HUMAN CAPITAL, AND PRODUCTIVITY IN FRANCE:

A GENERATIONAL ACCOUNTING PERSPECTIVE

by Xavier Chojnicki* AND Paul Eliot Rabesandratana

LEM–CNRS (UMR 9221), University of Lille

The aim of this paper is to highlight the potential productivity gains resulting from improvements in the (i) educational attainment and (ii) health status of the working-age population. For that purpose, we develop a Generational Accounting Model applied to the French economy. Using the conventional methodology of generational accounting, we first estimate the adjustments that will be necessary to ensure the sustainability of French fiscal policy in the long term under the assumption that individual taxes and transfers grow at the same rate as labor productivity. However, this assumption does not account for the explicit determinants of individual productivity. Therefore, we then explain how pro- ductivity growth is partly due to the French population�s skill level and its health level, which is approximated by the survival rate of adults. We estimate that the increased educational attainment and improved adult survival rate in France generate potentially important productivity gains that could significantly challenge the weight of the burden induced by aging. Therefore, we estimate that this change could reduce the tax burden bequeathed to future generations by 79 percent. Our results are robust to the main assumptions.

JEL Codes: E62, H51, I10

Keywords: generational accounting, health status, education, productivity, aging

1. Introduction

The French population is engaged in an aging process that, far from being specific to France, affects virtually all populations worldwide. Although the French population has aged considerably over the past 35 years—the median age increased from 31.7 to 40 years—the mass retirement of the baby boomers is con- siderably amplifying the effects of the aging process. According to INSEE1 (Blan- pain and Chardon, 2010), the total French population should increase by approximately 20 percent between 2007 and 2060, although the French labor force is expected to remain relatively stable. During the same period, the share of the population aged 65 and over should nearly double. These demographic evolu- tions would induce a significant increase in the old-age dependency ratio—that is, the ratio of the population aged 65 and over to the population aged 15–64

Note: This paper benefited from the valuable comments of H. d�Albis and L. Ragot. We thank the participants of the 2nd Lille–Ghent Workshop in Economics (Lille, 2013), the International Pen- sion Workshop (Paris, 2013), the Forum for Economists International (Amsterdam, 2013), the con- gress of the AFSE (Aix-en-Provence, 2013), and the internal seminar of the University of Lille for helpful comments and discussions. The usual disclaimer applies.

*Correspondence to: Xavier Chojnicki, LEM–CNRS (UMR 9221), Universit�e Lille 3, Maison de la recherche, Domaine, Universitaire du Pont de Bois, B.P. 60149, 59653 Villeneuve d�Ascq, France ([email protected]).

1The National Institute of Statistics and Economic Studies.

1

Review of Income and Wealth

DOI: 10.1111/roiw.12306

bs_bs_banner

872

Series 64, Number 4, December 2018

VC 2017 International Association for Research in Income and Wealth

years—in the coming decades. Having been approximately 25 percent in 2007, this ratio could be as high as 47 percent in 2060.

The impact of these developments on public finances is a source of con- cern.2 Based on current policies, public spending directly related to age could increase by approximately 3.1 percentage points of gross domestic product (GDP) according to the European Commission (2012). Most of this increase would come from spending on pensions (10.5 point), health (11.4 point), and long-term care (12.1 points). Potential savings in expenditures targeting younger segments of the population (spending on education and unemploy- ment in particular) would be very small (–0.4 for the former and 20.6 for the latter). Unfortunately, the slow growth of the French working-age population will not be able to absorb this increase in social spending. Therefore, the implicit public debt—that is, the present value of future deficits arising from population aging due to the rising costs of pensions, healthcare, and other age- related government expenditures—will substantially exceed explicit public debt.3

The use of the implicit public debt is called for because of the increasing diffi- culty of using the budget deficit as a good indicator of economic policy. Although it is widely used, the deficit exhibits the double disadvantage of being easily manipulated and, especially, static, thereby encouraging short-term policies. Use of the deficit would account for some receipts without integrating their counter- parts in terms of future commitments. To incorporate long-term public commit- ments, Auerbach et al. (1991) developed generational accounting (GA) as an alternative to the budget deficit to assess the sustainability of fiscal policy in the long term. This approach builds on an intertemporal treatment of the government budget constraint: at any date, the present value of government purchases must be covered by the current net public wealth, the present value of the net taxes that will be paid by future generations over their entire lives. This technique makes it possible to evaluate, for the base year, the actual value of the net payments that current generations (i.e. those of which one or more members are still alive today) will pay to the state until the end of their lives. Based on the state�s long-term budget constraint, the technique then compares the net burden carried by those born in the base year (the only generation to be followed across its entire lifespan) with the net burden to be carried by the generations to come (those born after the base year).

Using balanced growth assumptions, the classical methodology of GA defines the generationally balanced growth policy as a situation in which individ- ual taxes and transfers increase at the same rate as labor productivity. However, existing GA exercises rely on very simple assumptions regarding the changes in labor productivity across generations. These changes are usually related to exoge- nous growth rates that have no explicit link to the human capital level held by each generation. However, the role of human capital in productivity is now well

2It should be noted here that we will only focus on the fiscal impact of aging. Thus, the non-monetary costs of aging are outside the scope of our study.

3The implicit public debt could range between 253 percent of GDP (Bovenberg and Van Ewijk, 2011) and 295 percent (IMF, 2012).

2873

Review of Income and Wealth, Series 64, Number 4, December 2018

VC 2017 International Association for Research in Income and Wealth

established (Mankiw et al., 1992). Not accounting for increases in human capital inevitably leads to an overestimation of the negative effects of aging on productiv- ity and the public budget (Chojnicki and Docquier, 2007). The purpose of this paper is to revisit the standard GA methodology by introducing skill heterogene- ity and productivity gains induced by the continuous increase in French life expectancy.

The concept of human capital, which was first highlighted by Schultz (1961) and Becker (1964), is the productive capacity acquired through the individual accumulation of general and specific knowledge and skills. It primarily consists of the education and health of workers. Increased educational attainment in succes- sive cohorts affects the growth rate of labor productivity and influences the aver- age age profile of taxpayers and recipients of transfers over time. Indeed, ceteris paribus, a higher educational level implies higher labor productivity, higher labor income and, at least, higher net tax payments. This is why the age profile of net taxes is highly dependent on educational attainment. Thus, by disaggregating the generational account of each generation by schooling level, Chojnicki and Doc- quier (2007) manage to highlight the positive effects of the first component of human capital, namely, education.

Health is the second component of human capital that positively affects labor productivity. Nelson and Phelps (1966) and Grossman (1972) were the first to demonstrate theoretically how better health improves labor efficiency and, con- sequently, labor productivity. Subsequent contributions by Bloom and Canning (2005), Weil (2007), Aghion et al. (2012), Barro (2013), and O�Mahony and Samek (2016) empirically validate these theoretical findings. In addition, regard- ing the effects of health on productivity in the long term, Ashraf et al. (2008) undertake quantitative simulations to assess the potential productivity gains gen- erated by future health improvements. They estimate the effects of exogenous health improvements on output per capita in developing countries and report that the effects are slightly positive.

In this paper, we demonstrate that it is crucial to include workers� skill heteroge- neity and health status when evaluating fiscal sustainability for three main reasons. First, the age profile of taxes and transfers is highly dependent on educational attain- ment. According to our estimates, the present value of taxes paid by an individual over his or her entire life amounts to e91,126 for low-skill, e136,958 for medium-skill, and e210,392 for high-skill workers. Second, the skill composition of the population has changed dramatically over time and is likely to change further in the future. For example, in 2008, 72 percent of the 80-year-olds in France had educational attainment below a baccalaureate4 level, 13 percent had educational attainment between the bac- calaureate level and a university undergraduate degree, and only 15 percent had a higher qualification. For the cohort aged 30, these figures were 35 percent, 24 percent, and 41 percent. Even if one assumes stability in the level of education of future young cohorts, the average education level of the population will continue to grow due to the gradual rise of young and educated people in the age pyramid. Third, if the increase in life expectancy is likely to substantially affect the financing of social protection

4The baccalaureate is a French academic qualification, which is usually obtained at the end of high school.

3874

Review of Income and Wealth, Series 64, Number 4, December 2018

© 2017 International Association for Research in Income and Wealth

systems, these gains will also affect the working-age population and, potentially, posi- tively affect worker productivity. For example, Bloom and Canning (2005) demon- strate that a 1-percentage-point increase in adult survival rates translates into a 2.8 percent increase in labor productivity. This is why we regard improved health as a determinant of the evolution of the growth of labor productivity in this paper.

The remainder of this paper is organized as follows. Section 2 provides the mathematical tools for GA with heterogeneous skill and health status. Section 3 discusses data issues and the calibration of net tax profiles by age and education. Section 4 presents the results of the conventional GA methodology (Auerbach et al., 1991) with a single representative agent within each generation and where the labor productivity growth rate is considered exogenous. This conventional methodology yields our baseline scenario. We find that aging could generate an intertemporal public liability (IPL) of e2,260 billion (129 percent of 2010 GDP). The French government should increase all taxes by 13.5 percent or decrease all benefits by 14.5 percent to ensure the sustainability of French fiscal policy in the long term. Section 5 considers methodological issues affecting GA with skill het- erogeneity and health status. First, compared with the baseline, we show that the total burden left for future generations is reduced by 62 percent when accounting for the future change in the skill structure of the French population. Second, we observe that the productivity gains generated by improving the health of the French population could reduce IPLs by 16 percent relative to the baseline. Fur- thermore, we find that the simultaneous improvement of the skill structure and the health status of the French population should generate productivity gains which can reduce IPLs by 79 percent relative to the baseline.

A sensitivity analysis is presented in Section 6. Our results are quite robust to discounting assumptions, skill premium forecasts, the educational level of future cohorts, and the elasticity of health with respect to productivity. Section 7 sum- marizes and concludes.

2. The GA Model with Human Capital

2.1. Basic Features

Generational accounting (GA) was first introduced by Auerbach et al. (1991) and is a meaningful way to evaluate the sustainability of a fiscal policy. The GA method relies on the notion of an intertemporal budget constraint that requires that all public expenditures must be financed by taxes. For the base year, this can be expressed as follows:

PVLt1PVFt5PVGt2Wt:(1)

According to equation (1), the French Government Budget is balanced in the long term when the present value of government purchases, PVGt, less public net wealth, Wt, equals the sum of the present value of net tax payments by living generations over the rest of their lives, PVLt, and the present value of net tax payments by future generations over the rest of their lives, PVFt. Wt constitutes the only directly

4875

Review of Income and Wealth, Series 64, Number 4, December 2018

VC 2017 International Association for Research in Income and Wealth

observable element. Traditionally, it is considered equal to the opposite of the national debt, leaving aside the government�s wealth, particularly government assets.

The present value of government purchases, PVGt, is the discounted sum of public expenditures:

PVGt5 X1 s5t

Gs ð11iÞs2t

;(2)

where Gs measures public consumption that is not age specific in year s and i is the interest rate. Total public expenditures, Gs, is assumed to evolve under the double influence of population growth and productivity growth, which is equiva- lent to requiring expenditures to evolve according to productivity:

Gs ps

5ð11cÞs2t Gt pt ;(3)

where c is the rate of productivity growth and pt is the total population size in year t.

2.2. Introducing Skill Heterogeneity

The present value of net tax payments by living generations can be obtained by summing the present value of the net taxes that these generations will pay to the government over the rest of their lives; that is, by summing the generational accounts of living cohorts. We distinguish three educational levels (L 5 low skills, M 5 medium skills, and H 5 high skills) and assume that each individual lives for a maximum of D years. The present value of payments by living generations, PVLt, can be written as follows:

PVLt5 XD

j50

nLj;tp L j;t1n

M j;t p

M j;t 1n

H j;tp

H j;t

� � ;(4)

where pXj;t is the size of type X population (X 5L; M; H) of age j at time t and n X j;t

measures the generational account of these agents. The generational account sums the value of net taxes to be paid by each type

of individual over the rest of his or her life:

nXj;t5 1

pXj;t

XD

k5j

hXk;t1k2j p X k;t1k2j

ð11iÞk2j ; j50; . . . ; D; X 5L; M; H;(5)

where hXk;t1k2j is the net tax payment by an agent of type X and age k at time t1k2j. In practice, pXk;t1k2j can be projected using demographic forecasts (includ- ing mortality and net immigration flows), data on schooling levels per age, and estimates of the educational attainment of young living generations after the com- pletion of their education.

5876

Review of Income and Wealth, Series 64, Number 4, December 2018

© 2017 International Association for Research in Income and Wealth

Assume that there exist q types of taxes and transfers in the economy with q51; . . . ; Q, and let s be a tax if s > 0 and a transfer if s < 0. Then, we have

hXk;t1k2j 5 XQ

q51

sX;qk;t1k2j; X 5L; M; H;(6)

in which sX;qk;t1k2j describes the tax (resp. transfer) profile for a tax (resp. transfer) of type q of an agent who belongs to cohort k at time t1k2j and has skill type X.

2.3. Health Status and Productivity

Most existing GA models assumes time invariance of the age distribution of taxes and transfers except for the rate of technical progress:

sX;qk;t1k2j115ð11cÞ3s X;q k;t1k2j; X 5L; M; H:(7)

The mechanical application of a uniform growth rate means that economic growth does not change the age profiles of taxes and transfers, and that the fruits of growth are shared equally among the different cohorts. However, an increase in the average endowment of human capital per worker is potentially able to gener- ate substantial fiscal gains. According to human capital theory, explicit and par- ticularly important determinants of individual productivity are education and health.

By introducing skill heterogeneity in GA (Section 2.2), we are able to account for the effects of education on the evolution of labor productivity and, therefore, on the evolution of the present and future generational accounts. To be able to consider the second explicit determinant of individual productivity, we assume a simple theoretical framework in line with Aghion et al. (2012). In period t1k2j, workers with skill X produce Y Xt1k2j due to the production tech- nology FðAt1k2j; H Xt1k2jÞ as follows:

F At1k2j; H X t1k2j

� � 5At1k2j H

X t1k2j

� �q :(8)

The production of workers with skill X then depends on the total factor produc- tivity, At1k2j , and the health of workers, H Xt1k2j . Intuitively, a higher level of health makes labor more productive and therefore increases the amount of effi- ciency labor in the economy. Assuming that FðAt1k2j; H Xt1k2jÞ is a Cobb–Douglas production function, q describes the rate of return of H Xt1k2j in equation (8). Let cXt1k2j; g

A t1k2j and g

h;X t1k2j denote the growth rates of Y

X t1k2j; At1k2j and H

X t1k2j ,

respectively. Using a Solow growth decomposition (Solow, 1957), we obtain

cXt1k2j 5g A t1k2j 1 q3g

h;X t1k2j

� � :(9)

According to equation (9), the rate of productivity growth for workers with skill X at t1k2j is simply the sum of the rate of total factor productivity growth,

6877

Review of Income and Wealth, Series 64, Number 4, December 2018

VC 2017 International Association for Research in Income and Wealth

(gAt1k2j ), and the health improvement of the worker of type X, ðg h;X t1k2jÞ, weighted

by q. Inserting equation (9) into equation (7), we introduce an explicit link between individual productivity and worker health.

2.4. An Assessment of the Sustainability of Fiscal Policy

Adding equation (9) into our GA framework, we can then assess the finan- cial sustainability of public policies. Given the present value of payments by living generations, PVLt, the present value of government purchases, PVGt, and the net public wealth, Wt, we can easily determine the present value of net contributions of future generations, PVFt, as the residual of the intertemporal budget constraint (equation (1)). However, if this indicator can be read as a burden/surplus bequeathed to future generations, it appears difficult to imagine carrying forward the adjustment only for future generations (i.e. those born after the reference year). One way to proceed is to compute the hypothetical generational accounts of future cohorts under current fiscal policy. Using the same reasoning as in equa- tions (4) and (5) above, we write:

PVF�t 5 X1

s5t11

XMin s2t21;D½ �

j50

hLj;sp L j;s1h

M j;s p

M j;s 1h

H j;sp

H j;s

ð11iÞs2t ;(10)

where PVF�t is the present value of net payments by future generations if current fiscal policy is unchanged. This hypothetical value can then be compared with the residual value, PVFt, computed from equation (1):

� if PVF�t 5PVFt, the policy is sustainable and there is no need to make a fiscal adjustment;

� if PVF�t > PVFt, the government budget is in surplus and benefits could be increased without increasing taxes; and

� if PVF�t < PVFt, the current policy is not sustainable: this implies that current policy must be adjusted to restore sustainability.

As in Raffelh€uschen (1999), the sustainability of fiscal policy can then be esti- mated using the concept of intertemporal public liabilities (the “Sustainability Gap” in Auerbach et al., 1991) to obtain a measure of generational imbalances arising from following current fiscal policy for an infinite horizon. This indicator, IPL, is the residual of the intertemporal budget constraint, if all generations, pres- ent and future, receive the same tax treatment:

IPL5PVGt2Wt2PVLt2PVF � t :(11)

This indicator assesses the extent of reforms designed to restore balance (IPL50). If the present policy is unsustainable, the obvious strategy is to adjust taxes and/ or transfers at some future date. In this paper, we use an adjustment method that covers all members of all generations. If a gap has to be financed (to cover a defi- cit), we compute the proportional adjustment in all taxes (or all transfers) required to balance the budget.

7878

Review of Income and Wealth, Series 64, Number 4, December 2018

© 2017 International Association for Research in Income and Wealth

Let us decompose the net taxes on all generations into two basic components, taxes and benefits: hXj;s5h

X T;j;s2h

X B;j;s. A time-invariant adjustment factor can be

applied to each of these components (gT for taxes and gB for benefits) to restore sustainability. We then apply these proportional changes to both living genera- tions (over the rest of their lifetimes) and future generations to balance the budget constraint. Our adjustment rule is summarized by the following set of equations:

PVLadjt 5 XD j50

XD

k5j

X X 5L;M;H

hXT;k;t1k2jð11gTÞ2h X B;k;t1k2jð12gBÞ

h i pXk;t1k2j

ð11iÞk2j ;

PVF adjt 5 X1

s5t11

XMin s2t21;D½ �

j50

X X 5L;M;H

hXT;j;sð11gTÞ2h X B;j;sð12gBÞ

h i pXj;s

ð11iÞs2t ;

PVGt5PVL adj t 1PVF

adj t 1Wt:

There is a continuum of pairs (gT; gB) restoring the balance. Two specific pairs are usually considered, one with gT 50 if the balance is achieved through cuts in transfers and one with gB50 if the balance is achieved through tax increases.

3. Data Issues and Assumptions

The collection of data is the preliminary stage of any longitudinal exercise.

3.1. Population Forecast

Our baseline scenario uses the intermediate population projection of INSEE provided by Blanpain and Chardon (2010). We retain the intermediate population projection for the period 2007–60:

� the life expectancy at birth rises from 77.2 and 84.2 years for men and women in 2007 to 86.0 and 91.1 years, respectively, in 2060; � the fertility rate remains at 1.98 children per woman between 2007 and

2015 and decreases to 1.95 children per woman after this period; � the net number of persons immigrating to France is equal to 100,000 per

year until 2060, which corresponds to the average net annual migration flows observed since the mid-2000s.

This scenario includes a sharp increase in the old-age dependency ratio (i.e. the number of people aged 651 as a percentage of the number of people aged 15–64). The ratio was 25.3 percent in 2007 and is expected to reach approximately 47 per- cent in 2060 (see Table 1).

GA requires population projections to a very distant horizon. This is neces- sary to evaluate the net payment from living generations until the end of their lives, the value of public expenditures, and the generational accounts of future generations indefinitely. The INSEE forecasts are thus extended until 2110, assuming that the mortality, fertility, and migration rates remain at their 2060

8879

Review of Income and Wealth, Series 64, Number 4, December 2018

VC 2017 International Association for Research in Income and Wealth

levels. Nonetheless, GA attributes little weight to the net payment of generations to a fairly distant horizon due to discounting effects.

3.2. Educational Attainment

Three educational levels are distinguished: low-skill workers are those who have an education level below a baccalaureate (LS); medium-skill workers are formed by those who have completed at most two years of education above the baccalaureate level (MS), and high-skill workers are those who have completed more than two years of study above the baccalaureate level (HS). We use data on the skill composition of living French cohorts in 2008 taken from the French pop- ulation census (INSEE, 2008) for the population aged 30 and over. We assume that the skill structure of future cohorts (i.e. those aged 30 after 2008) to be con- stant. Thus, the resulting projection can be considered relatively pessimistic.5

The evolution of the skill structure of the French population over time is described by Figure 1. In 2008, 42.8 percent of the population aged 30 had a high skill level, compared with 22.1 percent who had a medium skill level, and 35.1 percent who did not even have a baccalaureate. Among those aged 60, these shares were 19.7 percent, 13.5 percent, and 66.8 percent, respectively. The main changes in educational attainment (as measure by the share of each educational group in the total French population) occurred before 2008. However, even assuming that there is no more progress in education attainment for younger cohorts (those aged 30 after 2008), the skill structure of the population is none- theless affected even in the future due to the rise of the younger cohorts, which have a high educational level, in the age structure of the population.

3.3. Tax and Transfer Profiles by Age and Educational Attainment

Estimating the age profiles of taxes and benefits for a reference year is the basis for any longitudinal calculation. We consider the six main branches of social security expenditures, corresponding to the different risks defined by social security account- ing: (1) retirement, (2) health, (3) family, (4) unemployment, (5) housing, and (6) pov- erty/exclusion. To these social security expenditures, we add education expenses, which also correspon

Collepals.com Plagiarism Free Papers

Are you looking for custom essay writing service or even dissertation writing services? Just request for our write my paper service, and we'll match you with the best essay writer in your subject! With an exceptional team of professional academic experts in a wide range of subjects, we can guarantee you an unrivaled quality of custom-written papers.

Get ZERO PLAGIARISM, HUMAN WRITTEN ESSAYS

Why Hire Collepals.com writers to do your paper?

Quality- We are experienced and have access to ample research materials.

We write plagiarism Free Content

Confidential- We never share or sell your personal information to third parties.

Support-Chat with us today! We are always waiting to answer all your questions.