Is an unfunded social security system good or bad for growth? A theoretical analysis of social security systems financed by VAT

• Author: Noritaka Maebayashi
• DOI: http://doi.org/10.1111/jpet.12403
• Published date: 01 August 2020
• Date: 01 August 2020

J Public Econ Theory. 2020;22:1069–1104. wileyonlinelibrary.com/journal/jpet © 2019 Wiley Periodicals, Inc.

|

1069

Received: 4 January 2019

|

Accepted: 12 September 2019

DOI: 10.1111/jpet.12403

ORIGINAL ARTICLE

Is an unfunded social security system good or

bad for growth? A theoretical analysis of social

security systems financed by VAT

Noritaka Maebayashi

Faculty of Economics and Business

Administration, The University of

Kitakyushu, Kokura Minami‐ku,

Kitakyushu, Fukuoka, Japan

Correspondence

Noritaka Maebayashi, Faculty of

Economics and Business Administration,

The University of Kitakyushu, 4‐2‐1

Kitagata, Kokura Minami‐ku,

Kitakyushu, Fukuoka 802‐8577, Japan.

Email: non818mn@kitakyu-u.ac.jp

Funding information

Ministry of Education, Culture, Sports,

Science, and Technology of the Japanese

government (JSPS KAKENHI),

Grant/Award Number: 17K18032

Abstract

This study investigates how unfunded public pensions

financed by value addedtax (VAT), as discussed in Japan,

affect economic growth and whether payroll tax (PT) or

VAT is the more growth‐friendly tax structure for

financing public pensions. We examine these issues

using overlapping generations models with parental

altruism and find that a public pension system financed

by VAT may increase economic growth when bequests

are operative. By contrast, when bequests are inoperative,

public pensions hinder growth unless agents are suffi-

ciently patient. Finally, public pensions financed by VAT

are more growth‐friendly than those financed by PT.

1

|

INTRODUCTION

Population aging driven by increasing longevity and low‐fertility rates has been making it

difficult to sustain unfunded social security systems.

1

In addition to population aging, many

countries have been suffering from lower growth and weak fiscal conditions, both of which

place pressure on the financing of pension systems. In these situations, many OECD countries

implemented important public pension reforms between 2013 and 2015 to improve the financial

stability of their pension systems. According to OECD (2015), Australia, Canada, Finland,

France, Ireland, the Netherlands, New Zealand, and Sweden have all raised payroll tax (PT)

rates to generate increased revenue for financing public pensions. In contrast to these countries,

1

OECD (2015) states as follows: “The share of individuals aged 65 and above will increase from 8% of the total world population in 2015 to almost 18% by 2050

and from 16% to 27% in the OECD. In the OECD, the share of the population older than 75years will be similar in 2050 to the share older than 65 years today.

Ageing directly affects the financing of pay‐as‐you‐go (PAYG) pension schemes, as a decreasing number of working‐age people has to sustain pension levels for

an increasing number of elderly.”

Japan has increased the consumption tax (value added tax, VAT) rate from 5% to 8% to improve

the financial stability of its social security systems, including public pensions, termed the VAT‐

public pension reform hereafter.

As discussed by Arnold (2008), when countries consider reforming their tax systems,

identifying the growth implications of different tax instruments aids the design of fiscal policies.

This is also the case with the policy design of unfunded public pensions. In particular,

policymakers that face difficulties financing public pensions must recognize the growth effects

of tax hikes through PT and VAT.

However, little scholarly attention has been paid to the difference in the growth effects

between a public pension financed by PT (a PT‐public pension hereafter) and a VAT‐public

pension. One exception is the study by Naqib and Stollery (1985), who investigate the effects of

two kinds of public pension financing on capital accumulation in a lifecycle model and show

that both a PT‐public pension and a VAT‐public pension reduce capital accumulation. The

negative effects of public pensions on capital accumulation are attributable to the decision‐

making of selfish individuals. Unfunded public pensions deter savings by selfish individuals

both through their tax burden and through their benefits (e.g., de la Croix & Michel, 2002).

Building on Naqib and Stollery (1985), this study considers the role of parental altruism to

examine the difference in the growth effects between PT‐and VAT‐public pensions. With

parental altruism, public pension benefits increase transfers to children (educational

investment and bequests) and can stimulate capital accumulation if the positive effect of

public pensions on altruistic transfers dominates the negative tax burden effects. To examine

this, we construct endogenous growth models with overlapping generations (OLGs) that have

the following three features. First, human capital accumulation associated with parental

altruism drives economic growth. Second, altruistic parents face a trade‐off between leaving

bequests and investing in their children’s human capital taking account of these relative

returns. Finally, public pension benefits are financed by VAT and PT.

In this study, we consider two kinds of parental altruism. One is warm‐glow altruism under

which agents experience the warm‐glow of giving to their closest children through either

education or bequests. The other is dynastic altruism under which agents care about all future

descendants when they decide on their educational spending and bequests.

With both types of altruism, the difference in financing the public pension affects economic

growth through the trade‐off of altruistic transfers between education and bequests. Under the

PT‐public pension system, PT lowers returns from educational investment (wage rate) and has

negative effects on growth. Conversely, under the VAT‐public pension system, an increase in

VAT substitutes consumption when young into educational spending for children, whereas it

substitutes consumption when aged into leaving bequests. The influence of the VAT‐public

pension on growth thus depends on how it affects relative returns between education and

bequests through these substitution effects.

Because we focus on the role of parental altruism to examine the growth effects of public

pension systems, the work of Lambrecht, Michel, and Vidal (2005) is relevant to this study.

Lambrecht et al. (2005) investigate the growth effect of PT‐public pensions in the warm‐glow

altruism model, finding that when bequests are operative, a PT‐public pension always hinders

economic growth, whereas when bequests are inoperative, a PT‐public pension cannot be

positive for growth unless individuals are sufficiently patient.

In contrast to the work of Lambrecht et al. (2005), we consider the growth effects of VAT‐

public pensions with two kinds of altruism (warm‐glow altruism and dynastic altruism). More

specifically, this study addresses the following research questions. First, is a VAT‐public

1070

|

MAEBAYASHI

pension good or bad for growth? Second, is a VAT‐public pension more growth‐friendly than a

PT‐public pension?

The main results from our study are summarized as follows. In the warm‐glow altruism

model, we obtain the following results. First, when bequests are operative, a VAT‐public

pension increases growth unambiguously if individuals’utility function is log‐linear. Even

under the constant relative risk aversion (CRRA) and constant elasticity of substitution (CES)

utility functions, a VAT‐funded public pension can be good for growth under certain conditions.

These results contrast to the negative growth effect of a PT‐public pension when bequests are

operative found by Lambrecht et al. (2005).

Second, when bequests are inoperative, a VAT‐public pension does not always increase

growth. It is not good for growth unless individuals are sufficiently patient. This is qualitatively

similar to the case of the PT‐public pension when bequests are operative presented by Lambrecht

et al. (2005). However, comparing the growth effects of the two kinds of public pension systems,

we find that a VAT‐public pension is more growth‐friendly than a PT‐public pension.

Further, in the dynastic altruism model, we find that a VAT‐public pension is neutral to

growth, whereas a PT‐public pension is bad for growth.

These results lead to the following implications. First, whether a VAT‐public pension is good

for growth depends on the country’s type of altruism. If parents have warm‐glow altruism and

are sufficiently altruistic to educate their children and leave bequests, introducing a VAT‐public

pension may enhance economic growth. By contrast, if parents are not sufficiently altruistic

such that bequests are inoperative, the large burdens of VAT‐public pensions may be bad for

growth. If parents have dynastic altruism, a VAT‐public pension is neutral to growth. Second, a

VAT‐public pension is more growth‐friendly than a PT‐public pension.

The key mechanisms that explain how a VAT‐public pension can promote economic growth

and be growth‐friendly are as follows. As mentioned above, VAT generates a substitution effect

between educational spending and consumption in youth and between leaving bequests and

consumption in old age. When bequests are operative, these two substitution effects offset each

other because of the trade‐off between leaving bequests and educational investment. Therefore,

VAT has no distortionary effect on education. This is different from the case of a PT‐public

pension under which PT lowers returns from educational investment (wage rate) and has a

negative effect on growth. In addition, VAT does not distort saving because it becomes neutral

to the intertemporal decision of consumption. Thus, tax can be neutral to both educational

spending and saving. Furthermore, VAT‐public pension benefits are partly transferred to

children as a bequest, which increases the disposable income of children and promotes saving,

educational spending, and economic growth.

Finally, this study also addresses the welfare implications. In the warm‐glow altruism model,

we obtain the following results. First, when bequests are operative, a VAT‐public pension

increases social welfare, whereas a PT‐public pension decreases welfare. Second, when bequests

are inoperative, both a VAT‐public pension and a PT‐public pension decrease social welfare.

However, the negative welfare effects of a VAT‐public pension are smaller than those of a PT‐

public pension. In the dynastic altruism model, a VAT‐public pension is neutral to social

welfare, whereas a PT‐public pension decreases social welfare.

1.1

|

Related literature

As mentioned above, Naqib and Stollery (1985) show that both VAT‐and PT‐public pensions

reduce capital accumulation in a selfish agent model. They also show that the negative effect of

MAEBAYASHI

|

1071