Ecological Economics 36 2001 311 – 326
ANALYSIS
The effects of ageing and an environmental trust fund in an overlapping generations model on carbon emission
reductions
Reyer Gerlagh
a,
, B.C.C. van der Zwaan
a,b
a
Vrije Uni6ersieit Amsterdam, Institute for En6ironmental Studies, De Boelelaan
1115
,
1081
HV, Amsterdam, The Netherlands
b
Stanford Uni6ersity, Center for International Security and Cooperation, Encina Hall, Stanford, CA
94305
-
6165
, USA Received 3 January 2000; received in revised form 30 May 2000; accepted 25 July 2000
Abstract
Most currently employed Integrated Assessment Models for climate change are of a dynastic nature. They employ the Ramsey rule linking the interest rate, economic growth, and the rate of pure time preference. This paper argues
that, although a dynastic framework might be convenient for economic analysis, it is restrictive and can be misleading. Overlapping Generations models, which do not employ the Ramsey rule, are more flexible and may give
results different from those derived from dynastic models. With the Integrated Assessment Model ALICE 2.0, it is shown how various assumptions on demographic change and public institutions can affect the interest rate, thereby
influencing the efficient greenhouse gas emission reductions. It is concluded that dynastic Integrated Assessment Models are in many respects inappropriate for providing policy makers with quantitative figures about the costs of
carbon dioxide emissions and their desirable reduction levels. © 2001 Elsevier Science B.V. All rights reserved.
Keywords
:
Resource conservation; Discounting; OLG models; IAMs; Ageing; Trust funds JEL classification
:
C68; J19; H43; Q28; Q48 www.elsevier.comlocateecolecon
1. Introduction
Recently, environmental
and resource
economists have increasingly been considering large scales of time. Problems such as a potential
climate change, the exhaustion of natural re- sources, or the safe disposition of long-lived ra-
dioactive nuclear
waste, compel
current generations to take into account to a high degree
what the results are of their actions in the far future. In the context of a large time horizon, the
discount rate determines the outcome of economic
Corresponding author. + 31-20-4449555; fax: + 31-20- 4449553.
E-mail addresses
:
reyer.gerlaghivm.vu.nl R. Gerlagh, bob.van.der.zwaanivm.vu.nl B.C.C. van der Zwaan.
0921-800901 - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S0921-80090000232-9
analysis to a larger extent than in research with respect to short-term questions.
This paper draws attention of the increased relevance of accounting properly for the discount
rate in economic modelling. In particular, this will be done with respect to the phenomenon of global
warming, which since the summits of Rio and Kyoto has become high on the agendas of both
politicians and scientists. Since a potential climate change caused by anthropogenic greenhouse gas
GHG emissions has been recognised as a major environmental problem, a number of so-called
Integrated Assessment Models IAMs have been developed to advise policy makers on the costs of
current GHG emissions and on the desirable GHG emission reduction levels. From these IAMs, one
can derive strategies with respect to the efficient levels of greenhouse gas concentrations in the
atmosphere, and estimate the expected global ef- fects of a ‘business as usual’ carbon dioxide emis-
sion pattern. Among the score of IAMs developed since the beginning of the 1990s, some of the more
well known are CETA Peck and Teisberg, 1992, DICE Nordhaus, 1994, MERGE Manne and
Richels, 1995 , MERGE-II Manne et al., 1995 and RICE Nordhaus and Yang, 1996.
Most IAMs presently used are of a dynastic type.
1
They assume, some more and others less explicitly, a central planner maximising a welfare
measure which aggregates utilities over different generations. The allocation of consumer goods and
environmental resources is calculated from the rules for welfare distribution over these genera-
tions. Alternatively, in a so-called overlapping generations OLG framework, the dynamic alloca-
tion is primarily determined by the distribution of property rights over generations.
2
The analysis of many dynastic models has led to intensive debate
on the kind of welfare functions to be used see, for example, Azar and Sterner, 1996. In this paper, it
is argued that the dynastic mechanism of welfare distribution exhibits characteristics that are unreal-
istic, sometimes even misleading. Its prevalent use in IAMs has led to unnecessary controversies
between researchers advocating efficient resource use, on the one hand, and those who advocate
resource protection, on the other hand.
The dynastic IAMs commonly use a constant Rate of Pure Time Preference RPTP to discount
future welfare levels before aggregation into a net present welfare measure.
3
A low RPTP implies that future benefits and costs are valued relatively
highly, whereas the use of a high RPTP means that relatively little value is attached to future benefits
and costs. The discount rate considerably affects the calculation of the emission costs and the
corresponding efficient carbon dioxide emission levels. Because of the importance of the role played
by the discount factor in economic modelling, the precise value to be taken is subject of invariable
controversy.
Some economists suggest that historic data on interest rates can be used to calibrate the level of
the RPTP used in IAMs, while these in turn determine the efficient environmental resource use
Nordhaus, 1994, Chapter 6. In their opinion, time preference should be based on past evolution of the
human use of resources. In other words, they assume that historic data determine which environ-
mental resource use is economically feasible. Others strongly oppose this so-called ‘descrip-
3
For a more general welfare aggregation, which employs a so-called aggregator function that has no constant RPTP, see
Koopmans 1960 and Koopmans et al. 1964. The distinction between a constant and variable RPTP is not without impor-
tance. Lucas and Stokey 1984, as well as Epstein 1987, point out that if one presumes a constant RPTP, the steady
states of the economy constitute a continuum. They state that this is an unrealistic characteristic of long-term economic
behaviour. It is a direct result of the assumption of time-addi- tivity, which in their opinion is a mere artefact of economic
modelling.
1
Dynastic models are also referred to as ILA infinitely lived agent or Ramsey models. We favour the term ‘dynastic’
since it has a natural connotation to the empathy between generations that form a dynasty. It includes OLG models with
empathy, since they have the same analytical properties as pure Ramsey models see e.g. Barro, 1974.
2
OLG models exist with dynastic properties, as mentioned in footnote 1. Generally, however, OLG models do not assume
interpersonal comparability see Roberts, 1980, for a discus- sion in relation to public choice theory and have rather
different dynamic properties Geanakoplos and Polemar- chakis, 1991; Kehoe, 1991.
-
tive’ view, which normally leads to using relatively high values of the discount rate, and argue for a
‘prescriptive’ approach. Here, one uses a low dis- count rate since future generations should not be
discriminated against by lower welfare weights see, for example, Broome, 1992. The rationale
behind this is that if discounting is too high, it is normally considered economically unfeasible to
sustain a high quality environment.
In practice, it proves difficult to bridge the gap between the ‘descriptors’ and ‘prescriptors’. In
this paper, it is argued that part of the disagree- ment between their opposing views is due to the
fact that both use a dynastic framework. Such a dynastic approach gives a convenient simplifica-
tion of the dynamic competitive equilibrium, but it treats discounting with insufficient reality. In a
dynastic model with constant RPTP, the interest rate is rigidly linked to economic growth via the
Ramsey rule, r : r + gg, where r is the interest rate, r is the RPTP, 1g is the intertemporal
elasticity of substitution, and g is consumption growth see e.g. Cline, 1992. Using an OLG-
model allows for a more flexible relation between interest rate and economic growth since there is
no central planner and, consequently, there is no planner’s rate of pure time preference. Various
factors can result in changes in the interest rate over time, such as local or global demographic
changes,
4
or a modification of the regime designed to
preclude an
over-exploitation of
the environment.
These possible, or even likely, exogenous and endogenous changes render the extrapolation of
the overall trend of historic interest rates to the future unjustified, as well as the rigid dynastic
assumption inappropriate. Therefore, we suggest not employing a rigid discounting relation as is
common practise in dynastic models. To show that the more flexible OLG approach is feasible in
the IAM context, we present ALICE 2.0, an Integrated Assessment Model with Overlapping
Generations OLG-IAM applied to the issue of climate change. The model extends the usual
IAMs — including the OLG-IAMs by Stephan et al. 1997, Howarth 1998, and Manne 1999 —
in three respects: it contains a demographic transi- tion, it specifies environmental damages as a loss
of an environmental amenity associated with an environmental resource, and it specifies a transfer
mechanism that distributes the value of this re- source to consumers. Although all three elements
have been studied — to some extent at least — before, see e.g. Auerbach et al. 1989, Krautkrae-
mer 1985, and Howarth and Norgaard 1992, respectively, the novelty of our analysis lies in the
fact that we introduce these elements in one con- cise OLG-IAM with climate change. Vis-a`-vis the
intergenerational distribution of property rights over environmental resources, this paper follows
the recent theoretical literature that analyses re- source use in the context of OLG models
Howarth and Norgaard, 1992; Gerlagh and Keyzer forthcoming. Yet, distributing ‘climate
endowments’ over generations is to some extent premature, given the small number of environ-
mental goods over which property rights have been established so far, and given the slowness
with which the set of such goods is currently expanded. Practicable policy suggestions require
the use of a richer OLG model for a more elaborate discussion, see Farmer and Randall,
1997. Nonetheless, the numerical results demon- strate that future interest rates can readily de-
crease
below the
levels predicted
by the
descriptive approach in dynastic models. The paper proceeds as follows. Section 2 de-
scribes ALICE 2.0, designed for the analysis of greenhouse gas policies on a world economy level.
The model elaborates on ALICE 1.2, which is extensively described and analysed in Gerlagh and
van der Zwaan forthcoming. Section 3 specifies four scenarios with different assumptions regard-
ing future demographic change and regarding the distribution of ‘climate endowments’. Section 4
compares the numerical outcomes of the scenar- ios. Finally, Section 5 draws conclusions.
4
A global demographic change is likely to take place during the 21st century, as will be pointed out below. Also changing
social security policies could have a major impact on future interest rates, such as regimes implemented in response to the
ageing of the population.
2. Model specification for ALICE 2.0
