Abstract
This paper studies the optimal linear tax-transfer policy in an economy where agents differ in productivity and in genetic background and where longevity depends on health spending and genes. If agents internalize imperfectly the impact of health spending on longevity, the utilitarian optimum can be decentralized with type-specific lump-sum transfers and Pigouvian taxes correcting for agents’ myopia and for their misperception of health spending’s effects on the economy’s resources. The second-best problem is examined under linear taxation instruments. It may be optimal to tax health spending, especially under complementarity of genes and health spending in the production of longevity.
Similar content being viewed by others
Notes
Sources: The Human Mortality Database (2009).
Note that longevity differentials are not here due to a selective involvement in wars: Swedish women born in 1900 did not experience any world wars.
Note that, in the presence of addiction, whether mortality factors like obesity, alcoholism, and drug consumption are under the control of agents or not is questionable.
Data: World Health Organization Statistical Information System (2009). The sample includes 193 countries.
Environmental factors of longevity include the quality of lands (Kjellström 1986), of waters (Sartor and Rodia 1983), and of the air (Kinney and Ozkanyak 1991). Genetic diseases take various forms, such as the sickle-cell disease and the familial hypercholesterolemy (favoring heart attacks; Soliani and Lucchetti 2001).
That picture comes from the study of Herskind et al. (1996), which relies on a sample of 2,872 Danish twin pairs born between 1870 and 1900.
That assumption is supported by the large empirical literature on limited rationality in the context of health-affecting behavior (see O’Donoghe and Rabbin 2000). In particular, the widespread feeling of invulnerability of young people shown by Quadrel et al. (1993) can be regarded as some myopia, i.e., an ignorance of the effects of one’s actions.
Note that we are not dealing here with the idea that individuals might rationally adopt a high rate of discount. In that case, government’s intervention is highly questionable and has been labeled “old paternalism”.
Thus, the critical utility level for continuing existence is set to zero (see Broome 2004).
Our focus on longevity-enhancing spending has important consequences when interpreting the results of this study. In reality, various health spending, which have little relationship with longevity, exhibit a strong redistributive dimension, or affect the quality of life periods or productivity. Whether such spending should be subsidized lies outside the scope of this paper, which focuses on longevity-improving spending.
Thus, this study complements other papers, such as Zhang et al. (2006) and Pestieau et al. (2008), which analyze the optimal taxation policy in a dynamic framework, but without an explicit heterogeneity in longevity-enhancing characteristics. An exception is Ponthiere (2010), who studies lifestyle-based longevity in a dynamic model where lifestyles are transmitted vertically or obliqually across generations.
This expression presupposes no pure time preferences, as well as a utility from being dead normalized to zero.
Note that this formalization of myopia is formally equivalent to assuming some pessimism of agents, in the sense that, under α < 1, the perceived probability of survival is always inferior to the actual probability. While this constitutes a simplification, that modeling of myopia has the virtue of analytical conveniency.
We will discuss the implications of those assumptions throughout this paper.
The assumption of a perfect annuity market, which is rather strong, is made here for analytical convenience. Actually, this study would like to abstract from tractability difficulties raised by accidental bequests, which are examined by Cremer et al. (2007). Assuming a perfect annuity market for each risk class is one way to avoid those difficulties. Another way consists in assuming that the government taxes entirely the savings of the dead (see Section 4).
A concern for responsibility implies that one pays attention, to some extent, to the relation between the initial conditions in which agents are and their final positions. Here, the social planner has, as a unique objective, the maximization of the sum of utilities (which, under welfarism, are the unique relevant pieces of information for positions). That maximization problem is only constrained by survival functions and utility functions and, thus, does not pay a specific attention to how conditions and positions are related.
The difficulties raised by varying longevity include, among other things, the definition of a critical utility level for continuing existence, making the addition of a new life-period yielding that utility level neutral. In the following, we rely on classical utilitarianism, where that critical level is fixed to zero.
Note also that if α i = α < 1 ∀ i, it would follow that θ i = θ ∀ i.
Equivalently, we assume that R i = 1 ∀ i. The absence of an annuity market consists of another major second best feature of the framework studied in this section.
A rate below 100% would complicate the analysis significantly, as some agents would benefit from unintended bequests. This would be, to some extent, close to the assumption of imperfect annuity markets, discussed in Cremer et al. (2007). For simplicity, we shall abstract here from those difficulties by assuming full taxation of the saving of dead agents.
It is also assumed, for simplicity, that agents cannot borrow with public pensions as collateral.
Here again, we assume, for simplicity, that an interior solution exists. See Andersen and Bhattacharya (2008) on the savings of myopic agents under a PAYG system.
To simplify the presentation, we shall assume, in this section that θ = P = 0, so that changes in τ affect the demogrant T only.
Total differentiation of the aggregate households’ budget constraint \(\frac{\left( 1-\tau \right) ^{2}Ew^{2}}{2}+T=0\) leads to \(\left( 1-\tau \right) Ew^{2}d\tau =dT\).
On that technique, see Cremer et al. (2008). There are other ways to combine the first-order conditions. One way allows to get rid of the multiplier μ, but the expression obtained by following that alternative approach would have a more difficult interpretation.
For conveniency, we shall here delete the * superscripts, but we remain at the optimum.
We now assume that θ = T = 0.
Note that here again, these are not real substitution effects, as we consider aggregate compensation.
For the ease of presentation, we now assume here that τ = P = 0.
The term \(\bar{e}\) denotes the average level of health spending in the population.
Take the case of two agents 1 and 2 with w 2 > w 1 > 0 and ε 2 > ε 1 = 0, and assume \(\pi \left( \varepsilon _{i}+e_{i}\right)\). Assume further that \(\pi _{1}\left( 0\right) =0\), \(\pi _{1}^{\prime}\left( 0\right) =\infty ,\pi _{2}(\varepsilon _{2})=1,\) \(\pi _{2}^{\prime}(\varepsilon _{2})=0\). One expects e 1 > 0 and e 2 = 0.
If e and ε are substitutable, \(cov\left( u^{\prime }\left( x\right) ,e\right) >0\), and we have a positive or a negative tax on health.
Clearly, if only e can enhance longevity, more productive agents do not spend less on health than less productive agent, as there exists, under π ε = 0, no way to ‘compensate’ low health spending in longevity terms.
References
Andersen T, Bhattacharya J (2008) On myopia as rationale for social security. CESifo working paper 2401
Becker GS, Philipson T (1998) Old age longevity and mortality contingent claims. J Polit Econ 106:551–573
Bender J, Trautner C, Spraul M, Berger M (1998) Assessment of excess mortality in obesity. Am J Epidemiol 147(1):42–47
Bommier A (2005) Uncertain lifetime and intertemporal choice: risk aversion as a rationale for time discounting. Int Econ Rev 47(4):1223–1246
Bommier A, Leroux ML, Lozachmeur JM (2007a) Social security and differential mortality. Working paper, University of Toulouse
Bommier A, Leroux ML, Lozachmeur JM (2007b) Uncertain lifetime, redistribution and nonlinear pricing of annuities. Working paper, University of Toulouse
Broome J (2004) Weighing lives. Oxford University Press, New York
Christensen K, Johnson T, Vaupel J (2006) The quest for genetic determinants of human longevity: challenges and insights. Nat Rev Genet 7:436–448
Cremer H, Lozachmeur JM, Pestieau P (2007) Collective annuities and redistribution. CORE discussion paper 96
Cremer H, De Donder P, Maldonado D, Pestieau P (2008) Taxing sin goods and subsidizing health care. CORE discussion paper 31
Doll R, Hill B (1950) Smoking and carcinoma of the lung. Br Med J 2:739–747
Eeckhoudt L, Pestieau P (2008) Fear of ruin and longevity enhancing investment. Econ Lett 101:1–3
Herskind AM, McGue M, Holm NV, Sorensen T, Harvald B, Vaupel J (1996) The heritability of human longevity: a population-based study of 2872 Danish twin pairs born 1870-1900. Hum Genet 97(3):319–323
Kaplan GA, Seeman TE, Cohen RD, Knudsen LP, Guralnik J (1987) Mortality among the elderly in the Alameda county study: behavioral and demographic risk factors. Am J Public Health 77(3):307–312
Kinney P, Ozkanyak H (1991) Associations of daily mortality and air pollution in Los Angeles county. Environ Res 54:99–120
Kjellström T (1986) Itai-Itai disease. In: Fridberg L, Elinder CG, Kjellström T, Nordberg G (eds) Cadmium and health: a toxicological and epidemiological appraisal, volume II: effects and responses. CRC, Boca Raton, pp 257–290
O’Donoghe T, Rabbin M (2000) Risky behaviour among youths: some issues from behavioural economics. Working paper 285, University of California at Berkeley
Pestieau P, Ponthiere G, Sato M (2008) Longevity, health spending and Pay-as-you-Go pensions. FinanzArchiv 64(1):1–18
Peto R, Lopez AD, Boreham J, Thun M, Heath C (1992) Mortality from tobacco in developed countries: indirect estimation from natural vital statistics. Lancet 339:1268–1278
Poikolainen K, Escola J (1986) The effect of health services on mortality decline in death rates from amenable to non-amenable causes in Finland, 1969–1981. Lancet 1(8474):199–202
Ponthiere G (2010) Unequal longevities and lifestyles transmission. J Public Econ Theory 12(1):93–126
Quadrel MJ, Fischhoff B, Davis W (1993) Adolescent (in)vulnerability. Am Psychol 48:102–116
Sartor F, Rodia D (1983) Hardness of municipal waters and cardiovascular mortality in four small Belgian towns. In: Addulla M, Nair B (eds) International symposium. Health effects and interactions of essential and toxic elements, Lund, 13–18 June. Charwell-Bratt, Bromley
Soliani L, Lucchetti E (2001) Les facteurs génétiques de la mortalit é. In: Caselli G, Vallin J, Wunsch G (2001) Démographie: analyse et synthèse, volume III, les déterminants de la mortalité. Editions de l’INED, Paris
Solomon C, Manson JE (1997) Obesity and mortality: a review of the epidemiological data. Am J Clin Nutr 66(4):1044–1050
The Human Mortality Database (2009) University of California, Berkeley, USA, and Max Planck Institute for Demographic Research, Germany. Available at http://www.mortality.org
World Health Organization Statistical Information System (2009) Data available online at http://www.who.int/whosis/en/. Data retrieved on 3 Feb 2009
Zhang J, Zhang J, Leung MC (2006) Health investment, saving, and public policy. Can J Econ 39(1):68–93
Acknowledgements
The authors are most grateful to Dirk Van de gaer for his comments and suggestions on this paper. We would like also to thank two anonymous referees for their helpful remarks on this paper.
Author information
Authors and Affiliations
Corresponding author
Additional information
Responsible editor: Junsen Zhang
Rights and permissions
About this article
Cite this article
Leroux, ML., Pestieau, P. & Ponthiere, G. Optimal linear taxation under endogenous longevity. J Popul Econ 24, 213–237 (2011). https://doi.org/10.1007/s00148-010-0304-1
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00148-010-0304-1