Abstract
We analyze the optimal life-cycle education decision of a single atomistic individual and show that the standard result of part-time education and part-time work throughout the life-cycle holds only under very special and unrealistic assumptions. Once these assumptions are relaxed, different education strategies become optimal. These range from switching back and forth between work and education (educational leave) to full-time education at the beginning of life and full-time work when older (classical schooling). The resulting strategies for investing in education are better aligned with the observable pattern of educational investments over the life cycle. The particular path chosen by individuals then depends on other aspects, e.g., on the individual’s lifetime horizon or on nonlinearities in human capital accumulation with respect to time invested in education.
This chapter is dedicated to the memory of Arkadii V. Kryazhimskii, who passed away while working on it. He is deeply missed.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
See, for example, the Encyclopædia Britannica (2018); stating that “When the number of sellers is quite large, and each seller’s share of the market is so small that in practice he cannot, by changing his selling price or output, perceptibly influence the market share or income of any competing seller, economists speak of atomistic competition[. . . ]”.
- 2.
In Romer (1990), human capital is the central determinant of economic growth because it is the crucial input in the production of new knowledge. This idea has been formalized and further developed by Funke and Strulik (2000), Dalgaard and Kreiner (2001), Strulik (2005), Bucci (2008), Strulik et al. (2013), Prettner (2014), and Prettner and Strulik (2016). For other beneficial effects of education see Lee and Mason (2010), Venti (2015), and Mason et al. (2016).
- 3.
In the model, the feasibility of such a decomposition rests on the assumptions of the lack of disutility of work beyond the opportunity costs of investing in human capital and that capital markets are perfectly competitive.
- 4.
Note that we assume in Theorem 2.1 that a balanced growth path with positive human capital growth exists and obtain as a result that also the control u ∗(t) ≡ 1 for all t ∈ [0, ∞) without human capital growth is optimal. This is, at first glance, a contradiction. But the assumption does not rule out that there exists a balanced growth path without human capital growth. From a mathematical point of view, the theorem could be reformulated to start with “If an optimal singular control exists” but we chose our formulation as economically more intuitive.
- 5.
The derivation of Eq. (2.22) can be found in Appendix 2.
References
Acemoglu, D. (2009). Introduction to Modern Economic Growth. Princeton University Press.
Aseev, S. and Veliov, V. (2014). Needle variations in infinite-horizon optimal control. Variational and Optimal Control Problems on Unbounded Domains, G. Wolansky, A. J. Zaslavski (Eds). Contemporary Mathematics, 619: 1–17.
Barro, R. J. (1991). Economic growth in a cross section of countries. The Quarterly Journal of Economics, 106(2): 407–443.
Bella, G., Mattana, P., and Venturi, B. (2017). Shilnikov chaos in the Lucas model of endogenous growth. Journal of Economic Theory, 172: 451–477.
Ben-Porath, Y. (1967). The Production of Human Capital and the Life Cycle of Earnings. Journal of Political Economy, 75(4): 352–365.
Blinder, A. S. and Weiss, Y. (1976). Human capital and labor supply: A synthesis. Journal of Political Economy, 84(3): 449–472.
Bucci, A. (2008). Population growth in a model of economic growth with human capital accumulation and horizontal R&D. Journal of Macroeconomics, 30(3): 1124–1147.
Cervellati, M., and Sunde, U. (2013). Life expectancy, schooling, and lifetime labor supply: Theory and evidence revisited. Econometrica, 81: 2055–2086.
Cohen, D. and Soto, M. (2007). Growth and human capital: good data, good results. Journal of Economic Growth, 12: 51–76.
Cuaresma, J. C., Lutz, W., and Sanderson, W. (2014). Is the demographic dividend an education dividend? Demography, 51(1): 299.
Dalgaard, C. and Kreiner, C. (2001). Is declining productivity inevitable? Journal of Economic Growth, 6(3): 187–203.
de la Fuente, A. and Domenéch, R. (2006). Human capital in growth regressions: How much difference does data quality make? Journal of the European Economic Association, 4(1): 1–36.
Edle von Gaessler, A. and Ziesemer, T. (2016). Optimal education in times of ageing: The dependency ratio in the Uzawa-Lucas growth model. The Journal of the Economics of Ageing, 7: 125–142.
Encyclopædia Britannica (2018). Monopoly and competition. https://www.britannica.com/topic/monopoly-economics#ref920520 [Accessed on 12/05/2018].
Farmer, J. D. and Foley, D. (2009). The economy needs agent-based modeling. Nature, 460: 685–686.
Ghez, G. and Becker, G. S. (1975). A theory of the allocation of time and goods over the life cycle. In The Allocation of Time and Goods over the Life Cycle, pages 1–45. NBER.
Haley, W. J. (1976). Estimation of the earnings profile from optimal human capital accumulation. Econometrica, 44: 1223–1238.
Funke, M. and Strulik, H. (2000). On endogenous growth with physical capital, human capital and product variety. European Economic Review, 44: 491–515.
Hanushek, E. A. and Woessmann, L. (2012). Do better schools lead to more growth? Cognitive skills, economic outcomes, and causation. Journal of Economic Growth, 17: 267–321.
Hanushek, E. A. and Woessmann, L. (2015). The Knowledge Capital of Nations: Education and the Economics of Growth. The MIT Press.
Hazan, M. (2009). Longevity and lifetime labor supply: Evidence and implications. Econometrica, 77: 1829–1863.
Hazan, M. and Zoaby, H. (2006). Does Longevity Cause Growth? A Theoretical Critique. Journal of Economic Growth, 11: 363–376.
Kirman, A. P. (1992). Whom or What does the Representative Individual Represent? Journal of Economic Perspectives, 6: 117–136.
Lee, R. and Mason, A. (2010). Fertility, human capital, and economic growth over the demographic transition. European Journal of Population, 26(2): 159–182.
Lucas, R. (1988). On the mechanics of economic development. Journal of Monetary Economics, 22(1): 3–42.
Lutz, W., Cuaresma, J. C., and Sanderson, W. (2008). The demography of educational attainment and economic growth. Science, 319(5866): 1047–1048.
Mason, A., Lee, R., and Xue Jiang, J. (2016). Demographic dividends, human capital, and saving. The Journal of the Economics of Ageing, 7: 106–122.
Prettner, K. (2014). The non-monotonous impact of population growth on economic prosperity. Economics Letters, 124: 93–95.
Prettner, K. and Strulik, H. (2016). Technology, Trade, and Growth: The Role of Education. Macroeconomic Dynamics, 20(5): 1381–1394.
Romer, P. (1990). Endogenous technological change. Journal of Political Economy, 98(5): 71–102.
Ryder, H. E., Stafford, F. P., and Stephan, P. E. (1976). Labor, leisure and training over the life cycle. International Economic Review, 17(3): 651–74.
Sala-i-Martin, X. (1997). I just ran two million regressions. American Economic Review, 87(2): 178–183.
Sala-i-Martin, X. S., Doppelhofer, G., and Miller, R. (2004). Determinants of Long-Term Growth: A Bayesian Averaging of Classical Estimates (BACE) Approach. American Economic Review, 94(4):813–835.
Strulik, H. (2005). The role of human capital and population growth in R&D-based models of economic growth. Review of International Economics, 13(No. 1):129–145.
Strulik, H., Prettner, K., and Prskawetz, A. (2013). The past and future of knowledge-based growth. Journal of Economic Growth, 18: 411–437.
Strulik, H. and Werner, K. (2016). 50 is the new 30 – long-run trends of schooling and retirement explained by human aging. Journal of Economic Growth, 21: 165–187.
Uzawa, H. (1965). Optimum technological change in an aggregative model of economic growth. International Economic Review, 6(1): 18–31.
Venti, S. and Wise, D. A. (2015). The long reach of education: Early retirement. The Journal of the Economics of Ageing, 6: 133–148.
Acknowledgements
We would like to thank the editor who handled our submission, Alberto Bucci, as well as an anonymous referee and Uwe Sunde for valuable comments and suggestions.
Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendices
Appendix 1: Proofs
Proof of Lemma 2.1
Consider an arbitrary admissible control u(t) and its corresponding human capital path h(t) being a solution to Eq. (2.3). Let x(t) and z(t) be solutions to
and
Then, k(t) = x(t) + z(t) for all t ≥ 0. Note that (2.5) and (2.19) imply that
such that the No Ponzi Game condition given by Eq. (2.4) leads to
The state variable z(t) does not play any role for the optimization problem with objective function given by Eq. (2.1) and can be omitted in further considerations.
Consider the No Ponzi Game condition in Eq. (2.20). From Eq. (2.18) it follows that
and so the condition given by Eq. (2.20) requires that
This inequality implies that the set of admissible controls c(t) is the largest if the term \(\int _0^{\infty } e^{-r(s,0)} w(s) h(s) u(s) {\,\mathrm d} s\) takes the maximal possible value. Thus, maximizing the right hand side of Eq. (2.21) will allow maximizing utility as given by the objective function in Eq. (2.1). Therefore, the original problem given by Eqs. (2.1)–(2.4) is reducible to the two separate problems put forward above. □
Proof of Proposition 2.1
The statement of the proposition follows from Theorem 4.1 in Aseev and Veliov (2014): Assumption (A1) is fulfilled by the dynamics of the problem, and Assumption (A2) in Theorem 4.1 is satisfied because the objective function is finite for all admissible controls. For finite objective functions, the notion of weakly overtaking optimality and strong optimality coincide (cf. Definition 2.3 and 2.5 in Aseev and Veliov, 2014). Furthermore, the notion of local weak overtaking optimality is weaker than the property of weak overtaking optimality (cf. footnote 3 in Aseev and Veliov, 2014). Hence, any strongly optimal control is also locally weak overtaking optimal. Thus, the theorem can be applied for this problem and implies that the Maximum Principle holds for the optimal control and the adjoint function can be explicitly stated as in the proposition.
The derivation of functional forms in Proposition 2.1 is purely technical and given in Appendix 2. Note that in the result of Aseev and Veliov (2014) the adjoint multiplier of the objective function in the definition of the Hamiltonian is equal to 1, which simplifies the analysis compared to other results in the literature. □
Proof of Theorem 2.1
We divide the proof into three steps. First, we show that the existence of an optimal balanced growth path with positive human capital growth implies the existence of a singular solution. Then, we derive a necessary condition for an optimal solution to exist. Finally, we show that infinitely many optimal solutions exist.
The assumption of an optimal balanced growth path with positive human capital growth implies that, for some constant d > 0, it holds that \(d = \dot h(t)/h(t) = \chi [1-u^*(t)]\). Thus, u ∗(t) is constant almost everywhere, u ∗(t) ≡ 1 − d∕χ < 1, for almost all t ∈ [0, ∞). The boundary value u ∗(t) ≡ 0 has objective value zero in the education problem, which is clearly not optimal because controls that lead to a positive objective value exist. Therefore, 0 < u ∗(t) < 1 for almost all t ∈ [0, ∞), which proves that a singular optimal solution exists.
Next, we derive the necessary condition for a singular solution presented in Corollary 2.1 by using the optimality conditions of Proposition 2.1. Consider the switching function ζ(t), defined as
If a control is singular over the interval [t 1, t 2) ⊂ [0, ∞), then the switching function ζ(t) is zero almost everywhere, which implies
Furthermore, \(\dot {\zeta }(t)=0\) for t ∈ [t 1, t 2). Using the adjoint equation (2.13), the derivative of ζ(t) with respect to time is given by
The condition \(\dot {\zeta }(t)=0\) implies that the first condition in Corollary 2.1 given by Eq. (2.16) is fulfilled.
Note that the conditions ζ(t) = 0 and \(\dot {\zeta }(t)=0\) are independent of the control u(t). If these conditions are fulfilled, every control u(t) satisfies the necessary optimality conditions and therefore infinitely many solutions and infinitely many singular solutions exist. We now show that all of these solutions are optimal. Considering any control u(t), the objective value of the education problem is given byFootnote 5
The objective value is thus maximized and equal to w(0)h(0)∕χ for any control u(t) that satisfies Eq. (2.15). This equation is in turn satisfied by infinitely many solutions, which proves the claim of the theorem. By shifting the initial point in time it follows that, along every optimal control, discounted future income is always equal to the current level of human capital multiplied by w(t)∕χ. □
Proof of Theorem 2.2
If r(t) and w(t) are constant, it follows that r(t) ≡ χ for t ∈ [0, ∞), while the wage rate per unit of effective labor is undetermined, w(t) ≡ w > 0.
Consider the maximal achievable income I(t) := χk(t) + wh(t), which is equal to the income at time t under a constant interest rate and a constant wage rate if the agent works full time. Taking the derivative of I(t) with respect to time, we obtain
Consumption can be rewritten as
Inserting this expression into the equation for \(\dot I(t)\), we obtain that the maximal achievable income grows at a constant rate, independently of the control u(t),
Due to our assumption 0 < (χ − ρ)∕(χθ) < 1, there exists \(\tilde u \in (0,1)\) such that \(\dot h(t)/h(t) = \chi (1-\tilde u) = (\chi - \rho )/\theta \). Inserting this \(\tilde u\) and the expression for c ∗(t) into the equation for \(\dot k(t)\), we obtain
Due to Corollary 2.1, the control \(u(t)\equiv \tilde u\) is also optimal, which concludes our proof of the existence of a balanced growth path. □
Appendix 2: Derivations
1.1 Derivation of Eqs. (2.12) and (2.13)
According to Theorem 4.1 in Aseev and Veliov (2014), the adjoint is defined by ξ(t) = Z(t)I(t) with Z(t) being the fundamental matrix solution to the problem
and
The solution to the differential equation with some initial condition z(0) = z 0 is
Therefore,
Simplifying the expression results in Eq. (2.12).
1.2 Derivation of Eq. (2.22)
For a given control path u(t), human capital at time t is given by
Using this expression, we obtain
Integrating by parts, where the first square bracket is differentiated, while the second is integrated, results in
Since the wage satisfies Eq. (2.16), the integrand of the second term is zero. Equation (2.16) further implies that w(t) = w(0)e r(t, 0)e −χt, which leads to the first term being equal to
which is Eq. (2.22).
1.3 Derivation of Eq. (2.23)
Equation (2.23) is a representation of optimal consumption—using the notion of “maximal achievable income” I(t)—which is not common. Therefore, we sketch how to obtain this formulation from the known equation (2.11).
We start by applying a time shift: exchanging in Eq. (2.11) for \(c^*_0\) the integration variable t by s and then the 0 by t, we obtain
Since r(t) ≡ χ for all t ∈ [0, ∞), it follows that r(s, t) = χ(s − t) and the integral in the denominator can be solved:
The numerator x(t) is by definition equal to the future discounted income plus today’s assets,
where the integral is, for any optimal control, equal to w(t)h(t)∕χ, see Eq. (2.17) and the first paragraph of Appendix 2. Combining the equations above, we obtain Eq. (2.23):
Rights and permissions
Copyright information
© 2019 The Author(s)
About this chapter
Cite this chapter
Skritek, B., Cuaresma, J.C., Kryazhimskii, A.V., Prettner, K., Prskawetz, A., Rovenskaya, E. (2019). Optimal Life-Cycle Education Decisions of Atomistic Individuals. In: Bucci, A., Prettner, K., Prskawetz, A. (eds) Human Capital and Economic Growth. Palgrave Macmillan, Cham. https://doi.org/10.1007/978-3-030-21599-6_2
Download citation
DOI: https://doi.org/10.1007/978-3-030-21599-6_2
Published:
Publisher Name: Palgrave Macmillan, Cham
Print ISBN: 978-3-030-21598-9
Online ISBN: 978-3-030-21599-6
eBook Packages: Economics and FinanceEconomics and Finance (R0)