Age‐related taxation of bequests in the presence of a dependency risk

• Published date: 01 February 2022
• Author: Marie‐Louise Leroux,Pierre Pestieau
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/jpet.12539

J Public Econ Theory. 2022;24:92–119.wileyonlinelibrary.com/journal/jpet92

|

© 2021 Wiley Periodicals LLC

Received: 22 March 2021

|

Revised: 7 July 2021

|

Accepted: 9 July 2021

DOI: 10.1111/jpet.12539

ORIGINAL ARTICLE

Age‐related taxation of bequests in the

presence of a dependency risk

Marie‐Louise Leroux

1,2,3

|Pierre Pestieau

3,4,5

1

Département des Sciences

Economiques, ESG‐UQAM, Montreal,

Quebec, Canada

2

CESifo, Munich, Germany

3

CORE, UCL, Louvain‐la‐Neuve,

Belgium

4

University of Liege, Liege, Belgium

5

Toulouse School of Economics,

Toulouse, France

Correspondence

Marie‐Louise Leroux, Université du

Québec à Montréal, Département des

sciences économiques, Case postale 8888,

Succ. Centre Ville, Montréal, Québec

H3C 3P8, Canada.

Email: leroux.marie-louise@uqam.ca

Funding information

Fonds de Recherche du Québec ‐Société

et Culture (FRQSC) FdR/SCOR Chaire

“Marché des risques et création de

valeur”; Social Sciences and Humanities

Research Council (SSHRC)

Abstract

This paper studies the properties of the optimal taxes

on bequests when individuals differ in wage and in

their risks of mortality and old‐age dependance. Sur-

vival is positively correlated to income but dependency

is negatively correlated with it. The government cannot

distinguish between bequests motives, that is whether

bequests resulted from precautionary reasons or from

pure joy of giving reasons. Instead, it observes the

timing of bequests and the health status at death.

Under the utilitarian social welfare criterion, we show

that bequests taxation results from a combination of

equity, insurance, and public revenue motives. If re-

distribution concerns dominate insurance concerns, it

is desirable to tax the most bequests of those in-

dividuals living long in good health and to tax the least

bequests of those dying early. This is a direct con-

sequence of the socio‐demographic structure we as-

sumed where richer agents live longer and in better

health than poorer agents. To the opposite, if insurance

concerns dominate redistributive concerns, early be-

quests should be the most taxed and, bequests under

dependency the least taxed. Under the Rawlsian cri-

terion, we find that early bequests should be the least

taxed and bequests left by the healthy long‐lived in-

dividuals should be the most taxed.

1|INTRODUCTION

In the recent years, the taxation of wealth transfers (including inheritance and inter vivos gifts)

has been going through a bad patch. Revenue raised through taxes on inherited wealth has

shrunk significantly over the last decades. In OECD countries, the proportion of total gov-

ernment revenues raised by such taxes has fallen from 1.1% in the mid‐1960s to 0.4% in 2017

(OECD, 2018). Over the same period, some countries like Australia, Israel, Mexico, Sweden,

and Norway have abolished death duties. One of the reasons for this state of disgrace may be

the poor design of those taxes (through exemptions, deductions, and avoidance opportunities).

The purpose of this paper is then to look at the optimal design of wealth transfer taxation, from

both an efficiency and a redistributive perspective, in a realistic framework where older in-

dividuals with different revenues face a double risk, the risk of death and that of dependence.

Following Cremer and Pestieau (2006), the optimal wealth transfer tax structure crucially

depends on the assumed bequest motives and, thus, on the type of bequests left by the de-

ceased. Some bequests are purely accidental. Since agents do not know how long they will live,

they may save more money than they turn out to need. In that situation, the taxation of

accidental bequests is quite efficient and can even be redistributive. However, savings can also

be motivated by the direct utility obtained from the act of giving (i.e., a “warm glow”moti-

vation). In that case, bequests are voluntary and reflect the preference of the deceased to leave

some money to their heirs so that inheritance taxation would end up being distortionary. Yet, it

is often quite difficult to disentangle those bequest motives. Instead, governments only have

information on the timing of bequests, that is, whether these are made early or late in life,

which in turn, provides some information regarding the motives of wealth transmission and on

which they can condition taxation.

Hence, savings comprise two main components. On the one hand, part of individuals'

savings aims at benefiting their heirs, out of altruism or joy of giving. On the other hand, part of

individuals' savings is precautionary and, aims at covering old‐age needs that cannot be (fully)

insured against. In case of a healthy retirement, these needs correspond to expenses not cov-

ered by pensions and due to annuity market imperfections. In case of old‐age dependence,

these needs correspond to the (extra) costs of long‐term care that are not insured. When agents

die prematurely or when they live long but healthy, precautionary savings are transmitted to

their heirs even though this transmission was not the main motivation. In most societies, the

risk of a too long life is better insured (for instance, through public or private annuitization)

than the risk of dependence.

1

In other words, one can expect that the precautionary savings

related to the risk of dependence represent an important part of those unplanned bequests.

2

So as to take into account these features, our paper models an economy where agents differ

with respect to their labor productivity, their risk of mortality, and their risk of dependence.

Those who happen to die prematurely, bequeath all their savings (both the “warm glow”and

the precautionary ones) to their children. Those who live long can be either healthy or de-

pendent. It is most likely that those who are lucky enough to remain autonomous can afford to

transfer also part of their precautionary savings to their children. To the opposite, dependent

agents will consume the entirety of their precautionary savings and, will only be able to

1

In most OECD countries, the market for LTC insurance is still largely underdeveloped (see, for instance, Brown &

Finkelstein, 2009; OECD, 2011; Pestieau & Ponthiere, 2011).

2

See Lockwood (2018) who shows that in presence of bequest motives, individuals prefer to make precautionary savings

rather than buying LTC insurance.

LEROUX AND PESTIEAU

|

93