Skip to main content

Precautionary public health, ageing and urban agglomeration

Abstract

Introducing precautionary public health policy into an overlapping generations model with migration of households between regions, we discuss the causality between ageing and urban agglomeration. We analyze the effect of public policy to expand longevity on an equilibrium population distribution between an urban region and a rural region in a steady state. As the result of our analysis, it is possible that the promotion of precautionary public health policy leads to enlarge wage differentials and enlarge regional disparity.

This is a preview of subscription content, access via your institution.

Fig. 1

(Source) OECD Health Statistics 2016 http://www.oecd.org/els/health-systems/health-data.htm. Ministry of Internal Affairs and Communications Japan, White Paper on Local Public Finance, 2012 http://www.soumu.go.jp/menu_seisaku/hakusyo/chihou/26data/2014data/26czb01-04.html

Fig. 2
Fig. 3

Notes

  1. Now we have not analyzed the causality between prevention expenditure and average life span age with econometric method in this paper. However, it is important to analyze these issues with econometric approach. Therefore, the analysis of this causality remains the issue of our future analysis.

  2. Ottaviano et al. (2002) develop a model to be able to derive the solution analytically using a quasi-linear utility function instead of the Cobb–Douglas utility function used by Krugman (1991) with a nonlinear general equilibrium system.

  3. Yakita (2011) extends a simple overlapping generation model by incorporating migration of households between regions and shows that the increase of spillover effects of regional public goods affects the population distribution in equilibrium. Introducing trans-boundary pollution caused by the manufactured goods sector in an urban area into Yakita (2011), Naito and Omori (2014) examine the effect of environmental policy on the equilibrium population distribution. Furthermore, introducing Marshallian externalities into Yakita (2011), Ikazaki (2014) shows the effect of social security system on equilibrium.

  4. Following the previous literature such as Yakita (2011) and Naito and Omori (2017) which assume the similar utility function, for the simplicity, we assume the utility function (1). However, when we assume the utility function including consumption at both periods, we can derive the similar results.

  5. Here we assume the competitive insurance market in this model. The saving of households is invested to the capital of production sector in region \(u\). Households who survive in the retirement generation receive interests and capital at the top of retirement generation.

  6. As for the interpretation of urban cost indexed by \(\sigma\), we follow Yakita (2011) and Naito and Omori (2014). We consider any congestion and traffic jam as urban cost. Urban cost may include not only congestion, traffic jams but also child-care costs. The child-care costs are not identical in both regions. We also assume that urban cost is constant for simplification though it is natural to assume that this urban cost depends on the population in region \(u\).

  7. See Appendix 2.

References

  • Becker GS, Barro RJ (1988) A reformulation of the economic theory of fertility. Q J Econ 103(1):1–25

    Article  Google Scholar 

  • Chakraborty S (2004) Endogenous lifetime and economic growth. J Econ Theory 116(1):119–137

    Article  Google Scholar 

  • Chen H-J (2010) Life expectancy, fertility, and educational investment. J Popul Econ 23(1):37–56

    Article  Google Scholar 

  • Diamond PA (1965) National debt in a neoclassical growth model. Am Econ Rev 55(5):1126–1150

    Google Scholar 

  • Ikazaki D (2014) Agglomeration, fertility, and economic growth. Stud Reg Sci 44(3):277–287

    Article  Google Scholar 

  • Krugman P (1991) Increasing returns and economic geography. J Polit Econ 99(3):483–499

    Article  Google Scholar 

  • Naito T, Omori T (2014) Can urban pollution shrink rural districts? Lett Spat Resour Sci 7(2):73–83

    Article  Google Scholar 

  • Naito T, Omori T (2017) Aging and urban agglomeration under a multi-regional overlapping-generations model. Rev Urban Reg Dev Stud 29(2):135–150

    Article  Google Scholar 

  • Ottaviano GI, Tabuchi T, Thisse JF (2002) Agglomeration and trade revisited. Int Econ Rev 43(2):409–436

    Article  Google Scholar 

  • Sato Y (2007) Economic geography, fertility and migration. J Urban Econ 61(2):372–387

    Article  Google Scholar 

  • Sato Y, Yamamoto K (2005) Population concentration, urbanization, and demographic transition. J Urban Econ 58(1):45–61

    Article  Google Scholar 

  • Yakita A (2001) Uncertain lifetime, fertility and social security. J Popul Econ 14(4):635–640

    Article  Google Scholar 

  • Yakita S (2011) Regional public goods, migration, and growth. Lett Spat Resour Sci 4(2):129–138

    Article  Google Scholar 

Download references

Acknowledgements

The previous version of this paper was presented in 54th annual meeting in Japan Section of RSAI and 7th Asian Regional Science Seminar. We are grateful to Osamu Keida, Hikaru Ogawa, Daisuke Nakamura, and anonymous referees for useful comments and suggestions. All errors remained in the paper are our responsibility.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Tohru Naito.

Ethics declarations

Funding

Funding was provided by Japan Society for the Promotion of Science (Grant nos. 15K03464, 16K12374, 16K03719).

Appendices

Appendix 1

Here, we will focus on the equilibrium value of \(G\).

From (21), we obtain

$$G_{t} \frac{{[\gamma + (1 - \gamma )p(G_{t} )]}}{{(1 - \gamma )p(G_{t} )}} = \tau Ab\left[ {\left( {(1 - \sigma )^{{\frac{ - \gamma }{{(1 - \gamma )p(G_{t} )}}}} - 1} \right)\theta k_{t} + 1} \right]$$
(34)

Here, let us assume that \(P/G_{t} > P^{{\prime }} (G_t)\). 1% increase in \(G_t\) produces increment of \(p\) less than 1%. In this case, the left-hand side of this equation increases with \(G_t\).

On the other hand, the right-hand side of the equation decreases with \(G_t\) because \(\theta^{{\prime }} (G_t) < 0\).

In this case, \(G_{t}\) is determined uniquely in equilibrium. It is easy to show that \(G_{t}\) is an increasing function of \(\tau\).

Appendix 2: The effect of \(\varvec{\tau}\) on steady state under agglomeration

First of all, we define \(\varOmega\) as follows.

$$\varOmega \equiv \left( {\frac{Az(1 - \alpha )}{{\gamma (1 - \sigma )^{\alpha } }}} \right)$$

Thus, we rewrite \(\Delta\) as follows:

$$\Delta \equiv \varOmega (1 - \tau )[\gamma + (1 - \gamma )p]^{\alpha } \cdot [(1 - \gamma p)]^{1 - \alpha } k^{*} - k^{*}$$

Differentiating \(\Delta\) with respect to \(\tau\), \(\frac{\partial \Delta }{\partial \tau }\) is given by

$$\begin{aligned} & \frac{\partial \Delta }{\partial \tau } = \varOmega \left\{ { - [\gamma + (1 - \gamma )p]^{\alpha } [(1 - \gamma )p]^{1 - \alpha } k^{*} } \right. \\ & + (1 - \tau )\alpha \left[ {(1 - \alpha )\frac{{{\text{d}}p}}{{{\text{d}}G_{t} }}\frac{{\partial G_{t} }}{\partial \tau }} \right][\gamma + (1 - \gamma )p]^{\alpha } [(1 - \gamma )p]^{1 - \alpha } k^{*} \\ & + \left. {(1 - \tau )[\gamma + (1 - \gamma )p]^{\alpha } (1 - \alpha )\left[ {(1 - \gamma )\frac{{{\text{d}}p}}{{{\text{d}}G_{t} }}\frac{{\partial G_{t} }}{\partial \tau }} \right]k^{*} } \right\}\frac{ > }{ < }0. \\ \end{aligned}$$
(35)

Moreover, we differentiate \(\Delta\) with respect to \(k^{*}\)

$$\begin{aligned} & \frac{\partial \Delta }{{\partial k^{*} }} = \varOmega (1 - \tau )\left\{ {\alpha \left[ {(1 - \gamma )\left( {\frac{{{\text{d}}p}}{{{\text{d}}G_{t} }}\frac{{{\text{d}}G_{t} }}{{{\text{d}}k^{*} }}} \right)} \right][\gamma + (1 - \gamma )p]^{\alpha - 1} [(1 - \gamma )p]^{1 - \alpha } k^{*} } \right. \\ & + \left. {[\gamma + (1 - \gamma )p]^{\alpha } (1 - \alpha )\left[ {(1 - \gamma )\frac{{{\text{d}}p}}{{{\text{d}}G_{t} }}\frac{{\partial G_{t} }}{{\partial k^{*} }}} \right][(1 - \gamma )p]^{ - \alpha } k^{*} ] - 1} \right\}\frac{ > }{ < }0 \\ \end{aligned}$$
(36)

Taking account of the signs of (35) and (36), the sign of \(\frac{{{\text{d}}k^{*} }}{{{\text{d}}\tau }}\) is not determined uniquely.

Since we define \(k^{**}\) as capital accumulation at steady state where 45° line and (28) intersect, we define \(\varGamma\) as the L.H.S of (33). Differentiating \(\varGamma\) with respect to \(\tau\), \(\frac{\partial \varGamma }{\partial \tau }\) is given by

$$ \begin{aligned} & \frac{\partial \varGamma }{\partial \tau } = \frac{\varLambda }{{\{ \gamma (1 - \sigma \theta k^{**} )\}^{2} }}\left[ {\gamma (1 - \sigma \theta k^{**} )\left\{ { - \tau p\left[ {1 + \theta k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} - 1} \right]} \right.} \right] \\ & + (1 - \tau )\left[ {1 + \theta k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} - 1} \right]\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{\partial \tau } \\ & + (1 - \tau )p\left[ {\left( {\frac{\partial \theta }{\partial \eta }\frac{\partial \eta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{\partial \tau } + \frac{\partial \theta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{\partial \tau }} \right)(1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} } \right. \\ & + \theta k^{**} \frac{1}{{\left. {\left. {(1 - \gamma )p^{2} } \right\}{ \ln } (1 - \sigma )^{{1 - \frac{\gamma }{(1 - \gamma )p}}} } \right]}} \\ & + (1 - \tau )p\left[ {1 + \theta k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} - 1} \right] \\ & \times \gamma \sigma k^{**} \left. {\left( {\left( {\frac{\partial \theta }{\partial \eta }\frac{\partial \eta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{\partial \tau } + \frac{\partial \theta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{\partial \tau }} \right)} \right)} \right]\frac{ > }{ < }0, \\ \end{aligned} $$
(37)

where \(\Delta\) denotes \(\frac{Abz(1 - \gamma )}{\gamma }\). Differentiating \(\varGamma\) with respect to \(k^{**}\), \(\frac{\partial \varGamma }{{\partial k^{**} }}\) is given by

$$\begin{aligned} \frac{\partial \varGamma }{{\partial k^{**} }} = \frac{\varLambda (1 - \tau )}{{\{ \gamma (1 - \sigma \theta k^{**} )\}^{2} }}\left[ {\gamma (1 - \sigma \theta k^{**} )\left\{ {\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{{\partial k^{**} }}} \right.\left[ {1 + \theta k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} - 1} \right]} \right. \hfill \\ \quad + p\left[ {k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} \left( {\frac{\partial \theta }{\partial \eta }\frac{\partial \eta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{{\partial k^{**} }} + \frac{\partial \theta }{\partial p}\frac{\partial p}{{\partial G_{t} }}\frac{{\partial G_{t} }}{{\partial k^{**} }}} \right)} \right. \hfill \\ \left.\left. {\quad + \theta (1 - \sigma )^{{^{{\frac{ - \gamma }{(1 - \gamma )p}}} }} + \theta k^{**} \frac{1}{{(1 - \gamma )p^{2} }}\ln (1 - \sigma )^{{1 - \frac{\gamma }{(1 - \gamma )p}}} } \right]\right\} \hfill \\ \quad \left. { + p\left[ {1 + \theta k^{**} (1 - \sigma )^{{\frac{ - \gamma }{(1 - \gamma )p}}} - 1} \right]\left( {\gamma \sigma \left( {\frac{\partial \theta }{{\partial k^{**} }}k^{**} + \sigma \theta } \right)} \right) } \right] - 1\frac{ > }{ < }0 \hfill \\ \end{aligned}$$
(38)

Taking account of (37) and (38), the sign of \(\frac{{{\text{d}}k^{**} }}{{{\text{d}}\tau }}\) is also not determined uniquely.

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Naito, T., Ikazaki, D. & Omori, T. Precautionary public health, ageing and urban agglomeration. Asia-Pac J Reg Sci 1, 655–669 (2017). https://doi.org/10.1007/s41685-017-0056-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s41685-017-0056-y

Keywords

JEL Classification