For the first course week, please read the article by Hsiang and Kopp (2018) and section 6 of the
Abridged Version of the Dasgupta Review. (For more detail, see section 4.2 and 4.3 of the Full
Report. I further recommend the Preface and Section 1 of the Abridged Version). The second
course week will mostly be based on chapter 4* of the Full Report (note that this is the
chapter after chaper 4).
There are therefore four avenues available to humanity for transforming the Impact Inequality
into an Impact Equality. They involve finding ways to:
(i)
reduce per capita global consumption;40
(ii)
lower future global population from what it is today;
(iii) increase the efficiency with which the biosphere’s supply of goods and services
are converted into global output and returned to the biosphere as waste; and
(iv)
invest in Nature through conservation and restoration to increase our stock of
Nature and its regenerative rate.
There are, however, limits to the extent our global demand can be reduced by being more
efficient in our consumption of goods and services (see Section 12 on technological
advances). Which is why our crude estimates say we must also invest in Nature and
attend to two problems that are rarely addressed in the economics of growth and climate
change: finding ways to reduce global per capita consumption (the required redistribution
measures would be enormous) and hastening the demographic transition in countries and
regions where larger families are the norm
We denote humanity’s global impact on the biosphere by I and the biosphere’s regenerative
rate by G. Both I and G are estimated in terms of the accounting values of the biosphere’s
goods and services, expressed, say, in real (international) dollars per year. I does not have to
equal G, for, continuing to use the parable of fisheries, the difference between the two would
automatically be accommodated by a change in the biosphere’s stock. We write the latter as
S. Declines in S are to be read as deterioration of the health of the biosphere. For vividness,
we could imagine that every ecosystem is valued at its accounting price, and S is the sum of
the accounting values of all the ecosystems on Earth. S is a stock, expressed in real
(international) dollars. As with the example of fisheries, G depends on S, so we write that
dependence as G(S). But as S is bounded, so is G bounded. N denotes global population, y
global GDP per capita, and α a numerical measure of the efficiency with which we are able to
convert the biosphere’s goods and services into the final products we produce and consume.
Global GDP is then N times y, which we write as Ny. The larger is α, the smaller is the
demand we make on Nature for the same level of GDP.44 Conversely, the larger is Ny, the
larger is the demand we make on Nature for the same level of α. We may then express I as
Ny/α and thereby write the fundamental inequality at the core of the economics of
biodiversity as
I = Ny/α > G(S)
Chapter 4 The bounded lobal economy
As our model has not yet been estimated, it cannot say when the global economy’s bounds
would bite so badly that further GDP growth would irreversibly tip the biosphere into a
different, and for us, worse, stability regime. Some might argue that boundedness of the
global economy could still leave room today for further GDP growth for years into the future
even while we invest to repair the biosphere. In theory there could be room, but the evidence
collated in the previous chapters runs counter to that possibility. Moreover, as is shown in
Chapter 13, by economic growth we should mean growth in an inclusive measure of wealth,
not GDP. Measuring economic prosperity by the size of GDP misleads hugely. The model
constructed here shapes the Review
Is GDP growth compatible with sustainable development? The question can be answered
only within the context of complete macroeconomic models of the long run, in which natural
capital plays an essential role – from source to sink. The model we construct here contains
those features and so can serve as a prototype of the kind governments and international
organisations should now construct. As the model economy is bounded, unbounded growth in
output, consumption and inclusive wealth is not possible. Nevertheless, one may ask whether,
while keeping consumption at politically acceptable levels it is possible for both GDP and
inclusive wealth to grow indefinitely even as they tend to finite limits. The answer is “yes”,
provided the stock of natural capital is large.
4.1 Substitutes and complements
Economic possibilities for the future depend on the extent to which goods and services are
substitutable among one another. Four kinds of substitution may be mentioned. First, there
can be substitution of one thing for another in consumption (e.g. nylon and rayon substituting
for cotton and wool; pulses substituting for meat). Second, produced capital can substitute for
labour and natural resources (e.g. the wheel in place of raw labour; refrigeration as way to
extend the durability of food). Third, novel production techniques can substitute for old ones
(e.g. the discovery of effective ways to replace the piston by the steam turbine). Fourth, and
for us here most importantly, natural resources themselves can substitute for one another (e.g.
wind and solar power in place of fossil fuels.) All this involves the more general idea that as a
resource is depleted, there are substitutes lying in wait, either at the same site or elsewhere,
meaning that even as constraints increasingly bite on any one resource base, humanity should
be able to move to other resource bases, either at the same site or elsewhere. The enormous
additions to the sources of industrial energy that have been realised (successively, human and
animal power, wind, timber, water, coal, oil and natural gas and, most recently, nuclear
power) are a prime historical illustration of that possibility.1
4.2 Modeling the global economy
The model we construct is highly stylised and can be disaggregated in obvious ways. We
assume that there is a produced good, whose output (Y) can be consumed (C) or invested
(I).162 Four categories of capital goods are required for producing Y: produced capital (K),
human capital (H), natural capital (S) – each of which is taken to be measurable as a scalar –
and publicly available knowledge (A, also a scalar). Natural capital, however, makes its
appearance in production in two forms: (i) as a flow of extracted provisioning service (R);
and (ii) as a stock supplying regulating and maintenance services (S) in the form of a global
public good. When referring to S we shall use the terms biosphere and natural capital
interchangeably
We are interested in studying economic possibilities in the biosphere’s safe zone. Rockström
et al. (2009) call the edges of that zone planetary boundaries (Section 4.1.2). If the biosphere
was allowed to cross that zone, it would fall into an unproductive state (Equation (4*.5)).
Global production Y is a power function of the four factors of production:
Y = ASβ Ka Hb R(1-a-b),
β > 0; a, b, (1-a-b) > 0
(4*.1)
Remark: But for the factor Sβ , which we explain presently, the production function in
equation (4*.1) is entirely orthodox. The combined factor Ka Hb R(1-a-b) is routine in
environmental and resource economics. Its functional form, with the exponents summing to 1
(i.e. production is subject to constant returns to scale in K, H, and R), is known as the CobbDouglas form, in honour of the economists who first gave prominence to it in their work on
agricultural production.163 The carrier of technological progress and institutional reform is
the factor A, which appears as a global public good. Growth economists refer to A as total
factor productivity because it can be regarded as the productivity of the comprehensive factor
of production Ka Hb R(1-a-b).
Our first move away from orthodoxy is to include a further global public good, denoted in
equation (4*.1) as S. The multiplicative factor Sβ (t), absent from nearly all contemporary
models, captures the fact that the human economy is embedded in the biosphere, like a family
sheltered in their home. As an example, if life in the oceans were to disappear, human life
would disappear (Chapters 1, 4).
R(t) is what we take from the biosphere directly (the provisioning services that yield fish,
useable water, timber, fibres, food), whereas Sβ (t) reflects regulating and maintenance
services (Chapter 2) – carbon and nitrogen cycles, disease control, climate regulation, soil
regeneration and so on – without which production, even life, would not be possible. We do
not ‘harvest’ regulating and maintenance services – their mere presence enables us to exist. In
contrast, the economy ‘draws on’ provisioning (and cultural) services. The distinction
between drawing upon Nature and dwelling in Nature is all important.
Time t ≥ 0 is continuous.
We express output at t as
Y(t) = A(t)Sβ (t)Ka (t)Hb (t)R(1-a-b)(t),
β > 0; a, b, (1-a-b) > 0
(4*.2)
Remark: The economy’s Total Factor Productivity (TFP) is A(t)Sβ (t). Even if A(t) recorded
an increase over a period of time, TFP would have declined if Sβ (t) had declined at a faster
rate. But the national income statistician would not know he had overestimated TFP if the
model he uses assumes β = 0. In Annex 3 to Chapter 13, we will show that in standard
national accounting practices TFP growth is overestimated: it is recorded as having been
large even if it were to have been negative. There is more than irony here: Under standard
accounting practices, the faster the global economy degrades the biosphere, the higher is
recorded growth in TFP.
The Impact Equation (Equation (4.4)) was constructed on the argument that to conserve is the
same, modulo sign, as to pollute. Human impact on the biosphere is R(t)+ Y(t)/αZ. We
imagine for simplicity that A plays a dual role, contributing positively to both Y and αZ. The
model is designed to provide the outlines of global economic possibilities on condition that
the biosphere remains within a safe zone for we humans. There are limits to the amount of
waste we can produce without tipping the biosphere into an uninhabitable stability regime.
That means there are limits to the extent to which αZ can be increased by changing our
practices and increasing A. We formalise this now
The efficiency with which output is converted into waste that does not unduly tax the
biosphere depends on both the state of knowledge (science and technology) and institutions.
Both are reflected in the parameter A. Thus αZ = αZ(A), which should be seen to be an
increasing function. But even if we imagine that continual investment in research and
development can raise A indefinitely, we have to acknowledge that αZ(A) is bounded above,
for otherwise, with sufficient human ingenuity Y/αZ could be made to go to zero for any
finite value of Y; and that would be to imagine we could eventually free ourselves of the
biosphere’s services entirely. Nor could αZ(A) go to infinity with Y at the same rate, for that
would be to imagine that the demands we make on the biosphere’s ability to absorb our waste
without experiencing breakdown is unbounded. If αZ(A) tends to infinity with A at a slower
rate than Y, then Y/αZ would tend to infinity with Y, and that would not be theoretically
possible on a bounded Earth. We may as well then suppose that αZ(A) is bounded above, by
say α*
As in Chapter 4, we assume heroically that the biosphere’s goods and services can be so
aggregated as to be measurable as a scalar quantity. So we suppose that the biosphere’s net
regeneration rate is a bounded function of S, G(S).
Thus
dS(t)/dt = G(S(t)) – R(t) – Y(t)/αZ,
α* ≥ αZ(A) ≥ αZ >0
(4*.3)
(R(t) + Y(t)/αZ) is humanity’s impact on the biosphere. Using equation (4*.1), we may define
αX by the equation R(t) = Y(t)/αX. 165 Any efficiency gain in extraction processes would be
subsumed in definition of A in equation (4*.1). Equation (4*.3) can then be expressed as
dS(t)/dt = G(S(t)) – Y(t)/αX – Y(t)/αZ,
αX > 0; α* ≥ αZ > 0
(4*.4)
Remark: Equation (4*.4) says that materials must balance: what is taken out of the biosphere
has to be put back in, albeit in a different (e.g. biodegradable) form (Section 4.7).166 The
global economy displays the Impact Inequality (Equation (4.3)) when dS(t)/dt < 0.
For concreteness, we return to the example in Box 3.3 and suppose
G(S) = rS[1 – S/S][(S – L)/S],
L, r, S > 0
(4*.5)
Equation (4*.4) then reads as
dS(t)/dt = rS(t)[1 – S(t)/S][(S(t) – L)/S] – Y(t)/αX – Y(t)/αZ,
(4*.6)
αX > 1; α* ≥ αZ > 0
Equations (4*.2) and (4*.6) say that if S was to fall below its safety zone (L), the human
economy would collapse eventually. We now investigate economic possibilities inside the
safety zone. The global population size is denoted as N(t). If human capital per capita is h(t),
H(t) = h(t) N(t). We take it that population is not controllable directly, but that future
population size can be influenced by investment in human capital. This influence is reflected
in the number of children desired by adults in a household. Call that number J. As evidence
that adults do have reproductive targets, Bongaarts and Cain (1981) noted that the rural
population in Bangladesh lost some 1.5 million children in the famine of 1974, and the
authors estimated that in the following year there were 1.5 million births in excess of what
would have been expected. Population size over time has a logistic shape (as in the model of
the biosphere in equation (4*.5)), so that
dN(t)/dt = N(t)[J(h) – N(t)],
J > 0, dJ(h)/dh < 0
(4*.7)
Remark: J is the long run population size (the logistic specification has been taken from
Arrow, Dasgupta, and Mäler (2003b)). That J is a declining function of h reflects the finding
that women’s education and knowledge of, and access to, modern family planning services
reduce desired family size. We could, should we wish, make J a function also of K and
assume that
(i)
∂J/∂K > 0 for small K (better diet, hygiene etc. encourages the birth rate to exceed the
death rate, perhaps even when adjusted for public investment in modern familyplanning services) and
(ii)
(ii) ∂J/∂K < 0 for large K (the cost of time increases with accumulation of produced
capital).
Remark: We could elaborate on equation (4*.7) by multiplying J(h) by a constant, say a,
whose value would reflect the extent to which the economy suffers from externalities, that is,
how far the desired number of children in a distorted economy differs from the desired
number of children in a non-distorted world. Chapter 9 discusses this issue in detail.
Investment, I, can take three forms: (i) accumulating produced capital K, which we write as I
K; (ii) accumulating human capital, which we write as I H; (iii) expenditure on research and
development so as to increase A, which we write as I A. National accounts then say
Y(t) = C(t) + I(t) = C(t) + I K(t) + I H(t) + I A(t)
(4*.8)
We assume produced capital depreciates at a constant proportional rate λ. That means net
investment in produced capital, dK(t)/dt, satisfies the condition
dK(t)/dt = I K(t) - λK(t)
(4*.9)
Combining equations (4*.8)-(4*.9) yields
dK(t)/dt = Y(t) – C(t) – I H(t) – I A(t) – λK(t)
(4*.10)
Equation (4*.10) records the net accumulation of produced capital.
Moreover, dA(t)/dt = I A(t)
(4*.11)
and dH(t)/dt = N(t)dh(t)/dt + h(t)dN(t)/dt
(4*.12)
Equations (4*.10)-(4*.12) represent net investment in produced capital, knowledge and
human capital. But in addition to the three forms of what one may call ‘active investment’
(Chapter 1), the economy admits a relatively passive form of investment, which can in
extreme cases involve simply ‘waiting’ (given a chance, forests regenerate, fisheries re-stock
and wetlands recover). This form of investment finds expression in equation (4*.6).
Equations (4*.2)-(4*.12) represent our model of the global economy when expressed in
stock/flow identities. In Chapters 11 and 13, we insert human institutions into it. But even the
stock/flow identities tell us that, as G(S) is a bounded function, Y must be bounded too
(Equation (4*.6)). We can buy time by reducing population growth and by raising the
efficiency parameters αX and αZ through technological change, institutional reforms,
substitution among factors of production, and wider behavioural changes; but no amount of
technological progress (A) and produced and human capital formation (K, H) can enable
global output Y to grow indefinitely, for sooner or later the biosphere will cross its safety
zone at L, and the human economy would in time cease to function. The idea of planetary
boundaries (Rockström et al. 2009; Steffen et al. 2015, 2018) puts scientific flesh into the
parameter L.
4.3 Contemprporary models of economic growth
Contemporary models of macroeconomic growth and development are extreme special cases
of our model, the extremities being that they imagine that boundedness of the biosphere does
not mean the human economy is bounded. The models assume implicitly that αZ can be made
to rise to infinity, at a minimum at the rate at which Y increases. We confirm that here.
Solow (1956): (i) A(t) = A (constant); (ii) β = 0; (iii) h = 1; (iv) 1-a-b = 0; (v) αZ =∞ ; (vi) J =
J(t) = n + N(t), n > 0.
In words, population grows exogenously at the rate n, and there is no technological progress,
no human capital accumulation, and no biosphere to constrain the economy. Y can grow to
infinity in the long run, but per capita output (Y(t)/N(t)) is bounded above.
Dasgupta and Heal (1974) and Solow (1974a): (i) A(t) = A (constant); (ii) β = 0; (iii) h = 1;
(iv) b = 0; (v) a > (1-a) > 0; (vi) λ = 0; (vii) αZ = ∞; (viii) r = 0; (ix) J = J(t) = N(t) = constant.
In words, population is constant and the biosphere is an exhaustible resource, which means it
is even more bounded than is envisaged in our model. The share of income attributable to
produced capital, K, exceeds the share attributable to the flow of natural resources in
production, R, that is, a > (1-a) > 0. There are sufficient substitution possibilities between
produced capital and exhaustible resources in production, so that even though R(t) has to tend
to 0 in the long run, Y is able to grow indefinitely to infinity, albeit at a rate that declines to
zero in the long run (Dasgupta and Heal, 1979).
Mirrlees (1967): This model is the same as Solow (1956), but for one difference, which is
that [dA(t)/dt]/A(t) = g > 0.
In words, the world economy enjoys a constant, exogenous rate of technological progress.
That allows not only Y to grow indefinitely, but allows per capita Y also to increase to
infinity.
Brock and Taylor (2010): The model assumes pollution is a by-product of output, but that
there is an abatement technology whose efficiency grows at a constant exogenous rate. The
model can be interpreted as assuming: (i) β = 0; (ii) 1-a-b = 0; (iii) [dA(t)/dt]/A(t) = g > 0;
(iv) J = J(t) = n + N(t), n > 0; (v) h = 1; (vi) [dαZ /dt]/αZ = m > 0, implying that α* = ∞.
It is shown by the authors that not only Y, but Y/N can grow to infinity.
Stiglitz (1974): The model makes two amendments to Dasgupta-Heal-Solow. (i)
[dA(t)/dt]/A(t) = g > 0; (ii) J – N = n > 0.
The author shows that technological progress can overcome the limitations imposed on an
economy by the exhaustibility of resources (oil and natural gas, coal) and can allow the world
to enjoy indefinite growth in Y. It is shown that if g > (1-a-b)n, per capita output can grow
indefinitely.
Remark: Romer (1986) and the many papers that followed it assume that technological
advances require investment. Which is why they are called ‘endogenous growth models’.
Despite the investment costs, indefinite growth in output is assumed to be possible. It follows
that, other things equal, growth possibilities in endogenous-growth models are more
circumscribed than growth possibilities in Mirrlees (1967), Stiglitz (1974) and Brock and
Taylor (2010), because in the latter three publications technical progress is assumed to occur
without cost at a positive exogenous rate.
For this week's S&G lecture I recommend reading chapters 4, 16, and 17 of the Abridged version of
the Dasgupta Review (which shall also be helpful for next week's guest lecture). In the full
Dasgupta Review, this week's material can mostly be found in chapter 13 (particularly Box
B13.2.1, Box 13.6, Annex A13.1).
I had the impression that some students had difficulty intuitively understanding some basics of
the first two course lectures. To those students, I recommend particularly reading the preface and
chapters 1, 5-7 of the abridged version. Chapters 2 and 3 of the abrigded version further give more
background about bioeconomic dynamics (depreciation and regeneration of nature). I hope this is
helpful.
4.3 The Impact Inequality
Ny/α +Ny/α , has exceeded aggregate supply G(S) per unit of time. That reads as XZ
The Impact Inequality as presented in expression (4.3) applies to the biosphere as a whole.
Although the notion of ecological footprint (the left-hand side of the Inequality) can be applied
to any group of individuals – from the individual and the household, to nations and the global
population – trade in commodities and services breaks the link between demand (Ny/α) and
supply (G(S)) for economic units smaller than the world as a whole. The ecological footprint of
a nation will not balance the regenerative rate of its ecosystems if its trade in the biosphere’s
goods and services does not balance, in units of biospheric material. Of course, it could be that a
country pays for its imports, perhaps even at their appropriate prices, but that is a different
matter. Here we are only formulating a way to break down the global imbalance of demand and
supply of those goods and services into imbalances among groups in the global population; we
are not discussing ‘fair trade’. Trade is discussed further in Chapter 15.
If the global ecological footprint, I, exceeds the biosphere’s regenerative rate, G, the biosphere as
a stock diminishes, and the gap between I and G increases. Similarly, if the footprint is less than
the biosphere’s regenerative rate, the stock increases, and the gap between I and G increase
without making additional demands on the biosphere provided either α or α and thus XZ
shrinks.140 However, either global population (N) or global output per capita (y), or both, could α
was to increase correspondingly. Improvements in technology (e.g. substituting degradable
waste for persistent pollutants; decarbonising the energy sector) and institutions and practices
(e.g. establishing Protected Areas; reducing food waste), and appropriate redistributions of
wealth are among the means by which α can be raised.
The factors affecting our demand for the biosphere’s goods and services, namely N, y, α and X α ,
affect one another. When a region of the Amazon rainforest is converted into cattle ranches, Z the
transformation would be expected to raise food production by raising the efficiency with which
land is used to grow crops (a rise in the corresponding α ), but it lowers α (industrial fertilisers
and pesticides degrade the soils and water bodies). The transformation could be read as
reducing S, or alternatively, because the composition of the biome changes, it could be read as a
less productive G-function. The overall effect would be to widen the gap between G(S) and
(Ny/α +Ny/α).
Abridget version
Chapter 4 Classifying and Valuing Assets
Assets are durable. Durability, of course, does not mean everlasting, but unlike services, assets are not
fleeting. It is tempting to call all assets capital goods, a term that has proved to be so attractive that it
now stretches to include knowledge (‘knowledge capital’); the law, the market system, and financial
institutions (‘institutional capital’); mutual trust, social norms, and group solidarity (‘social capital’);
culture and personal norms (‘cultural capital’); even religion (‘religious capital’). Economists have
been more reticent; they confine the use of the term to assets that are measurable.24 In the past,
economists reserved the term ‘capital goods’ even more stringently than they do now, for they only
included assets that are material (tangible) and alienable (i.e. whose ownership is transferable). Roads,
buildings, machines and ports are ready examples. As patents held by a firm are part of the firm’s
asset base, they appear in its balance sheet. So intangible and alienable assets are also included on the
list of capital goods. Taken together, they are called produced capital. The range of capital goods in
the economist’s lexicon has broadened over the years to include intangible but non-alienable assets
such as health, education, aptitude and skills, which, taken together, form human capital. Economists
today include human capital as a category of capital goods because they have discovered ways to
measure its value – not only to the individuals who acquire it, but also to society at large. In the past
decades, economists have developed methods for measuring the value individuals place on natural
resources; so we now have a third category of capital goods: natural capital. The methods can be
involved, for natural capital ranges over plants (they are tangible and alienable), pollinators (they are
tangible and often non-alienable), the view from one’s sea-front home (it is intangible and alienable)
and the global climate (it is intangible and non-alienable).
As the Review explores reasons for the growing disparity between private incentives and public
aspirations, we pay particular attention to the wedge between market prices of capital goods,
especially the market prices of natural capital, and what should be called their ‘social worth’, or
alternatively, their ‘social scarcity value’. Economists call them accounting prices. A capital good’s
accounting price is the contribution an additional unit of it would make to societal well-being (or
more narrowly, the common good).25 Being durable, capital goods offer their services over their
lifetime. So the accounting price of a capital good reflects the contribution its flow of services over its
lifetime makes to social well-being. Box 4 shows that absolute values of ecosystems (they are a form
of natural capital) have no meaning; only comparisons of values tell us something
What about knowledge, institutions and social capital – are they not assets as well? They are,
but as with biodiversity, measurement problems abound (try, for example, to compare the
value to a nation of ‘good governance’ with the real estate value of its capital city). So we
call them enabling assets, their worth is reflected in accounting prices. Enabling assets
bestow value to an economy’s capital goods. Even if the composition of capital goods is the
same in two societies, the one where people trust one another with reason would be found to
be wealthier than the one where people are distrustful of one another. Mutual trust is an
enabling asset: other things equal, the accounting price of, say, scientific journals in a country
at peace would be higher than in a country torn by civil strife. Likewise, other things equal, a
more biodiverse ecosystem is more productive in terms of the regulating and maintenance
services it supplies. So biodiversity is also an enabling asset.
The relationship runs the other way also. Knowledge in the sciences, technologies, and the
arts and humanities (they are all enabling assets) is created and acquired, but it is created and
acquired by people (human capital) in combination with produced capital (libraries,
laboratories) and natural capital (raw materials and ecosystem services). Institutions are also
created by people, as is mutual trust, which is the glue that holds communities together.
Financial capital facilitates exchange (among people and across time), so it too is an enabling
asset.
There are therefore three categories of capital goods in our classification and a wide range of
enabling assets. Enabling assets may not be measurable, but that does not matter: they enable
human societies to function well, and that can be measured. Accounting prices reflect that
worth.
16 Arbitraging Assets
Let us then imagine for concreteness that in their role as private asset managers people desire
to maximise their personal wealth and value stocks at their market prices, but that as citizens
they commend, even vote for, economic programmes that protect and promote societal wellbeing. The latter means that as citizen investors people value investment opportunities in terms
of accounting prices. Because markets place little value on Nature, the private investor’s
portfolio does not contain many investments that protect and promote Nature (that is, ‘green’
projects), while citizen investors favour portfolios that include green projects. Nevertheless,
whether people act as private investors or commend investment projects as citizen investors,
their goal is to maximise portfolio values – the big difference between the two types of investors
lies in the portfolios they believe maximise wealth. In the former case, the wealth in question is
private wealth; in the latter, the wealth in question is what we have previously called inclusive
wealth. The difference between them lies in the prices used to value capital goods. To have a
sharp distinction between the ‘private’ and ‘public’ roles we all play, let us imagine the citizen
investor is concerned with the global portfolio of assets.83 In either case, to say that an investor’s
goal is to maximise wealth is to say that her goal is to maximise the value of the portfolio from
among the portfolios available to her for consideration.
16.1 Arbitraging Assets at a Point in Time
The investor is interested in the real return, of course, but that is uncertain if only because she
cannot forecast changes in the consumer price index. That uncertainty will make itself felt
equally
for all assets. So we may ignore it in the account that follows and suppose that the return on
government bonds is certain.
In order to choose her portfolio, the private investor assesses the income she is likely to receive
from each of the stocks that are available for purchase. Income from a stock (i.e. the dividend
she would receive) is the stock’s yield. But the price of the stock relative to the government bond
could go up or down during the year.84 We call that capital gains, recognising that the ‘gains’
could be a loss. At the end of a year the investor’s wealth as capitalised in the stock would be its
yield plus the value of that stock. The latter will contain the capital gains term. Therefore, the
rate return on investment in a stock is the yield on a unit of that stock plus the capital gains on it
over the year.
In contrast, the interest on £1 on the government bond would be the bond’s yield. At the end of
the year, the investor’s wealth as capitalised in the bond would be the amount she receives
when she cashes in the bond. In other words, the rate of return on investment in the bond is the
interest on the bond over the year. Her wealth at the end of a year is her initial investment in the
portfolio of her choice plus the return she earns on it. It is then obvious that value maximisation
leads the private investor to choose a portfolio in which, adjusting for risk, the assets offer the
same rate of return. This rule is known as the arbitrage condition.85 And she holds a portfolio
rather than a single asset so as to reduce uncertainty in her wealth in a year’s time (as in the
saying that one should not put all of one’s eggs in the same basket).
16.2 Arbitraging Assets Across Time
In standard economic accounts, investment is interpreted as foregone consumption. That means
investment in a capital good is the increase in its quantity or improvement in its quality that will
be realised tomorrow if resources are diverted to it from consumption today.90 We are talking of
net investment, that is, investment net of depreciation. In the contrasting case of pollution,
depreciation is the decline in societal well-being caused by an increase in pollution (e.g.
depreciation of the atmosphere in its role as a sink for carbon emissions is the net increase in
carbon concentration).
Discussions on social discount rates on consumption have been prominent in cost benefit
analysis and the economics of climate change. The rates in use in policy assessment have almost
always been positive in sign. Why?
Three reasons suggest themselves: (i) the citizen investor may desire to award a lower weight
to a marginal unit of consumption in the future simply because it will be in the future; (ii) there
is a risk that humanity will cease to exist in the future; and (iii) the citizen investor expects the
average level of consumption to be higher in the future, meaning that a marginal addition to
consumption in the future will have less value to a future person than it has to someone today.
Reason (i) reflects myopia, (ii) reflects discounting for risk of extinction, and (iii) reflects a
desire to discount to compensate for intergenerational inequity. Notice though that reason (iii)
says that if the expectation is that people will be poorer in the future, then the consumption
discount rate could be negative.
There is a fourth, more subtle reason the citizen investor may choose to place a lower weight to
a unit of consumption of future generations relative to that of the present generation. It has to
do with countering a bias that the productivity of capital displays toward the future. Because
well-chosen investment has a positive yield, time has a direction in economic reasoning. A unit
of investment gives rise to more than a unit of additional consumption in the future. The Review
(Chapter 10) shows that unless the temporal asymmetry is countered, reasons (i)-(iii) may not
be sufficient to ensure a reasonably egalitarian distribution of well- being across the
generations.
The above reasoning shows that consumption discount rates reflect a combination of the citizen
investor’s ethical values and her reading of socio-ecological possibilities. The Review shows that
the discount rate (the rate could vary over time) she will want to deploy on future consumption
is the percentage rate at which the accounting price of consumption declines over the year in
question. The citizen investor’s choice of consumption discount rates has powerful implications
for her portfolio selection
17 Inclusive Wealth and Social Well-Being
There are two types of economic evaluation that we all conduct as citizens. One is policy
analysis, the other is sustainability assessment. To see what they involve, consider the list of six
questions we all ask frequently:
1. (1) How is the economy doing?
2. (2) How has it been doing in recent years?
3. (3) What would be our projection of the economy in the future if policies and
institutions evolve in the way we expect them to evolve?
4. (4) How is the economy likely to perform under alternative policies?
5. (5) Which policies should we support?
6. (6) What would be an ideal set of policies?
The questions prompt the citizen investor to make two types of comparisons:
Questions (1)-(3) require that she assesses whether an economy is on a path of sustainable
development; that is, she evaluates a change that has been or is likely to be experienced by an
economy as it moves through time. That is sustainability assessment. An example of the kind of
question she asks is: “Are the prospects of improvements in people’s lives and the lives of their
descendants better now than they were a decade ago?”
Questions (4)-(6) on the other hand prompt the citizen investor to engage in policy analysis.
There she evaluates the economic change that would be brought about by a proposed shift in
policy at a point in time. An example of the kind of question she asks in policy analysis is: “Is the
wetland preservation project being proposed likely to improve people’s lives and the lives of
their descendants?”
Question (3) is the projection of the socio-ecological world around which both sustainability
assessment and policy analysis are conducted.
It transpires that sustainability assessment and policy analysis, though seemingly different,
amount to the same exercise.95 It transpires also that the measure our citizen investor should use
to conduct the exercise is an inclusive measure of wealth, that is, adjusted for demographic
changes, the sum of the accounting values of produced capital, human capital and natural
capital. So we call the measure inclusive wealth (Figure 20).
Adjustments for demographic changes over time could be taken to mean inclusive wealth per
capita, rather than inclusive wealth itself. Figure 9, which presented the Managi-Kumar (2018)
estimates of movements in the three components of global inclusive wealth per capita in the
period 1992 to 2014, signalled just that.
Inclusive wealth is the measure of an economy’s productive capacity. If the inclusive wealth
per capita we bequeath to our descendants is greater than the inclusive wealth per capita
we ourselves inherited, we would be leaving behind a larger productive base for each of our
descendants. Being an aggregate figure, inclusive wealth does not reflect its distribution across
people. An enormous literature on distributive justice can be brought to bear to sharpen
sustainability assessment and policy analysis. In doing that though, the citizen investor would
be reading the distribution of inclusive wealth, not the distribution of GDP
Why should our citizen investor want economic accounts to include inclusive wealth? The
reason is, movements in inclusive wealth track well-being across the generations exactly. If a
change increases well-being across the generations, it records a rise in inclusive wealth.
Conversely, if
a change has an adverse effect on well-being across the generations, it records a decline in
inclusive wealth. Inclusive wealth and well-being across the generations are not the same
object, but they move in the exactly the same way.98
How can a measure that is designed to reflect an economy’s productive capacity also reflect
well-being across the generations? The answer is that in estimating inclusive wealth, accounting
prices are used to value capital goods. If we were now to return to the definition of accounting
prices (Section 5), we would recognise that they provide the link between wealth and wellbeing. Though it may seem materialistic to speak of wealth, in talking about inclusive wealth
rather than wealth without the qualification, we do not create any dissonance between our
ethical concerns and our concerns about the means to use for enabling lives to flourish. Which is
why it should cause no surprise that appraising investment projects amounts to examining
whether the projects contribute to inclusive wealth.
Economists have long advocated that the criterion for project appraisal should be the net
present value (NPV) of the flow of social benefits. The idea is to measure the flow of benefits, net
of costs, in terms of the accounting values of the flow of goods and services. The procedure then
involves summing the flow of net benefits, discounted at social discount rates. But summing a
project’s benefits over time amounts to the change in inclusive wealth that would be brought
about by the project. It is entirely satisfying that a criterion long in use in social cost- benefit
analysis matches the requirement that policy analysis should be conducted in terms of the effect
of policies on inclusive wealth. Notice though, there is no connection between GDP and the NPV
of investment projects. To advocate the use of GDP to measure economic progress while
advancing NPV as the criterion for project appraisal is bad economics. We have no explanation
for why the two have managed to survive simultaneously.
The citizen investor now understands that she has been misled all these years in being told that
GDP growth amounts to economic progress, that her concerns over biodiversity loss run against
what she is told is sound economic reasoning. She realises that the change from one period to the
next in inclusive wealth amounts to inclusive investment, that is, the accounting value of net
changes over the period in the quantities or qualities of all three classes of capital goods.
She now puts pen to paper to confirm that inclusive wealth increases from one period to the
next if (and only if) consumption in that period is less than net domestic product (NDP), that
is, GDP minus the accounting value of depreciation of all capital goods. She realises that it is
possible for GDP to increase over a period of time even as inclusive wealth increases. But she
also realises that because the Impact Inequality globally is so large today, the possibility of
global GDP growth generated by activities that cause depreciation of natural assets for an
indefinite period into the future, even while inclusive global wealth increases, is highly unlikely.
She realises that only formal economic models that include natural capital can, once they have
been estimated from data, provide an answer.
She may also feel that for equity reasons, raising material standards of living in low income
countries and regions ought to take precedence. And because she is persuaded that individual
preferences are socially embedded, she is sanguine that citizens of wealthy countries would not
be sacrificing anything of substance in changing their patterns of consumption to be
less damaging.
Our citizen investor realises as well that economics, when practised correctly, reflects her ethics
and provides her with a grammar for expressing her concerns – opposite to what she has been
told over the years. Once she has understood all that, she will complain to her government that
it is relying on a misleading index of economic success and is correspondingly mismanaging the
economy’s assets. She will ask her economic and finance ministry to explain, for example, why
its economists do not include the accounting value of natural capital in the investment projects
they appraise or in the public policies they advocate. She will ask her government to explain
to her why it is not doing more to repair a global financial system that permitted the world’s 50
largest banks last year to provide more than US$2.6 trillion in loans and other credit to sectors
that have an adverse impact on biodiversity (forestry and agriculture).99
She will then bring her complaints to the attention of her community. Her hope is that if citizens
act together, not only will her government listen, but the world will listen too. She will insist
that eternal vigilance may be the price we have to pay for freedom, but that that freedom will
prove to be meaningless if we are unable to prevent Nature from utter exhaustion.
Chapter 7 Human Institutions and Ecological Systems, 1: Unidirectional Externalities
and Regulatory Policies . Full review
Human activities involving the biosphere (in other words, all human activities) give rise to
externalities because property rights to large segments of the biosphere are either weakly
defined or inadequately enforced. And a common reason for the latter is that Nature is mobile.
No one can contain the atmosphere they befoul, the soil they contaminate, the rivers they
pollute. Moreover, the harms they cause are non-excludable
The connection between property rights and the distribution of wealth may seem obvious.
If one group of people in an economy owns few assets while another group owns a large
proportion of assets, the distribution of wealth in the economy is patently very unequal. In the
presence of externalities, though, the link between property rights and wealth distributions is
sufficiently subtle to go unappreciated. A distinction needs to be made between polluters’ rights
and pollutees’ rights. Continuing the example above, under the latter, inhabitants downstream
have a right to compensation from the logging company; under the former, inhabitants would
be obliged to pay the logging company to reduce its activity
Externalities and rights
Two categories of externalities are involved. One consists of the adverse externalities that are
conveyed through the material world because the Impact Inequality today is large (Chapter 4).
It is increasingly hard to argue that the vast quantities of produced capital and scientific and
technological knowledge we will be bequeathing future people compensate for the vastly
diminished biosphere we are leaving behind for them.
The other category consists of the externalities implicit in socially embedded preferences
(Chapter 9). Bauer’s critique does not acknowledge that individual households may themselves
affirm that their reproductive behaviour falls short of what they would ideally favour because
they are unable to coordinate their decisions with other households. As in every other field of
personal choice, it should be asked whether a collection of reasoned decisions at the individual
level harbour collective failure. This is the central question raised by externalities, and it is
particularly apposite in the case of adverse externalities and socially embedded preferences.
That family planning services bring many benefits (such as improved health, education, income,
and female empowerment) to those who make use of them has been documented repeatedly in
recent years (UNFPA, 2019). The focus in this Review on externalities points to the fact that
they bring benefits to others as well. Those additional benefits need to be included in the design
of social policies. In what follows, we study the tools that can be used for eliminating the
externalities that arise from our production and consumption of goods and services.
Reproductive externalities raise deeper and more difficult issues, which may be why they have
not been studied much in the literature on externalities. We present a few of the difficulties in
Chapter 9. The subject remains unsettled.
7.3 Taxing and Subsidising Externalities
A striking feature of well-functioning markets is that they make people responsible to others for
what they produce and consume. The qualifier ‘well-functioning’ is intentional and significant.
Well-functioning markets ensure that people pay the social cost of the resources they use, which
means that market prices correspond to accounting prices (Chapters 1 and 10). Well-functioning
markets harbour no externalities because they ‘internalise’ potential externalities.
Thus, if qi is the market price of asset i and ei is the value of the externalities generated by the
deployment of a marginal unit of i, then the asset’s accounting price is: pi =qi +ei
(7.1)
It is simplest to present the logic of Pigouvian taxes and subsidies by returning to the example
Pigou himself used – that of a firm whose factory emits toxic fumes. We assume that, other than
the harmful externalities the fumes give rise to, the economy does not suffer from any
distortion. The model is timeless.
Let p be the price at which the firm sells its product. The cost of producing Q units of output is
C(Q), where C(Q) = 0, dC(Q)/dQ > 0, and d2C(Q)/dQ2 > 0. In words, production cost
is zero if output is nil, and marginal cost of production is positive and increases with the volume
of production. In the absence of regulation, the firm’s profit function is [pQ – C(Q)], which we
write as φ(Q), as in Figure 7.2. That means the profit maximising output is the value of Q at
which the marginal cost of production equals the product’s price:
p = dC(Q)/dQ (B7.2.1) (Equation (B7.2.1) can also be expressed as dφ(Q)/dQ = 0).
Let Q be the solution. Q is the value of Q at the peak of the curve φ(Q).
Production creates toxic fumes as by-product. If the by-product matches output unit for
unit, we do not need a separate symbol for it. Denote the damage to residents in the
neighbourhood, expressed in monetary terms, by D(Q). Reasonably, we assume that
D(0) =0, dD(Q)/dQ > 0, and d2D(Q)/dQ2 > 0.
In words, damage is zero if production is nil and
the damage caused by a marginal unit of production is positive and increases with output. We
now introduce a regulator, whose objective is social well-being, V, which in our simple example
is the firm’s profit minus the damage caused by the firm’s production:
V(Q) = pQ – C(Q) – D(Q) (B7.2.2)
To make the idea of Pigouvian environmental taxes transparent, let us imagine that the
economy is subject to a ‘command and control’ system of authority, in that the regulator can
instruct the firm on how much to produce. Maximising V(Q) with respect to Q then yields the
socially optimal output level of production:
p = d[C(Q) + D(Q)]/dQ = dC(Q)/dQ + dD(Q)/dQ (B7.2.3)
Let Q* be the solution of equation (B7.2.3). The regulator would instruct the factory owner to
produce Q*. It is simple to confirm that Q* < Q, which is what intuition demands: output should
be reduced, as that is the only way to reduce pollution.
Notice that in deriving equation (B7.2.3) we have shifted the entire burden of agency from the
factory owner to the regulator. The Pigouvian solution to the problem of externalities is instead
to award agency over the choice of tax rate to the regulator but retain the agency
ver production decision to the factory owner. Formally, we have a two-stage game in which the
regulator makes the first move by imposing a tax rate, followed by the firm having to choose its
production level.
To see how the two-stage setting avoids command and control over production decisions,
suppose the regulator imposes a tax per unit of pollution, t*, equal to marginal damage at the
socially optimum level of production, dD(Q*)/dQ. Then the firm’s profit function, net of tax,
would be pQ – C(Q) – t*Q. Profit maximising output would then satisfy the condition that the
market price of the product equals the marginal cost of production plus the tax:
p = dC(Q)/dQ + t* = dC(Q)/dQ + dD(Q*)/dQ (B7.2.4)
It is simple to confirm that the solution of equation (B7.2.4) is Q*. Ideally, the optimal tax
revenue, t*Q* would be returned as a lump sum payment to the parties suffering damage. The
latter would enjoy a surplus, by the amount t*Q* - 0 Q*[D(Q)]dQ > 0. Notice that the regulator
knows in advance what the firm’s response would be for any choice of t. That means he can
ensure any level of output by a suitable choice of t. To take an example, imagine that the damage
caused by the pollution has a tipping point (Chapter 3), say Q ̃, in that pollution levels in excess
of Q ̃ are deemed to be unacceptable. The damage function in this case would display a steep
increase near Q ̃, in the extreme a large jump at Q .̃ It follows from equation (B7.2.4), the
optimum pollution tax ensures that Q* is less than Q ̃.
Pigouvian subsidies for beneficial externalities are obtained by a mere change of signs. Suppose,
as in the text below, a landowner converts a portion Q of his land into wilderness. Let D(Q) now
denote the benefit to neighbouring farmers from the wilding. Equation (B7.2.4) would now yield
the Pigouvian subsidy to the landowner per unit of land wilded.
Figure 7.2 Net Profit Function of Polluting Firm
Rationale for Environmental Regulations
The model analysed in Box 7.2 assumes the regulator knows both the damage function
D(Q) and the cost function C(Q). That is altogether too strong a requirement. The former
problem can be removed without much fuss. It could be, for example, that the regulator holds
reasoned beliefs about the harm factory smoke can inflict on people and that those beliefs can
be summarised as a (subjective) probability distribution over an uncertain damage function
(Chapter 5). In that case we would interpret D(Q) as the regulator’s expected damage function.
The problems that arise if the regulator does not know the firm’s cost function are less easy to
address. It may be reasonable to suppose that the factory owner knows the cost function (or at
least knows a lot more about it than the regulator). In contrast to the firm’s owner, who would
know how to respond to a tax, the regulator would be uncertain of the firm’s response. We
therefore have a setting in which the participants hold different kinds of information. The
regulator’s task is to make the best of it.
To formulate the regulator’s problem, let ɛ̃ be a random variable reflecting his ignorance
of the firm’s cost function. We may then represent his knowledge of the cost function
as C(Q, ɛ̃). Imagine now that the regulator wants to maximise the expected value of the Vfunction in equation (B7.2.2). By assumption the factory owner knows the true value of ɛ̃, but
the regulator does not. However, the regulator is able to calculate what the owner’s response
would be in every possible realisation of ɛ̃. That is because the regulator knows that if the
realisation of ɛ̃ is ε, then the owner’s profit function is pQ – C(Q, ε). This gives rise to an
interesting problem in incentives.
Suppose, as previously, that there is a tipping point Q ̃. With a Pigouvian tax, the regulator would
be able to ensure that the firm produces less than Q ̃ only by setting a high tax rate. But from the
regulator’s point of view (expected value of V), a high tax would lead to an undue contraction of
expected output. On the other hand, setting the tax rate too low would run the risk of an output
response from the owner that exceeds Q .̃ A balance has to be struck, and the way to do that is to
find a non-linear tax schedule, that is a tax rate per unit output, t(Q), that is not independent of
Q. For example, a natural compromise is a Pigouvian tax that comes tied to a condition that
output does not exceed a level that is chosen to be well below Q .̃ That is like a tax schedule in
which the tax rate increases sharply with output in the neighbourhood of the tipping point
If there are several (possibly many) firms polluting the atmosphere, a mixed tax-regulation
schedule suggests itself. The idea is to set a maximum for total emissions but then allow a quasimarket system to settle the emission allocation among polluters. A mechanism that
has been put into place in some parts of the world for containing global carbon emissions is
tradeable emission permits (Asian Development Bank, 2018; European Commission, 2020). The
number of permits issued equals the upper limit on aggregate emissions, expressed say in tons
of CO2 per year. Firms may emit only an amount equal to the number of permits it holds. One
variant would have the government sell the permits, the sales revenue serving as a tax the
government keeps. Another variant would be to have the government issue the permits to firms
and allow them to buy or sell permits in an open market. In either case the resulting market
prices for permits simulate Pigouvian taxes on global emissions, subject to an upper limit to
aggregate emissions.
Markets for extermanilities
Well-functioning markets do not harbour externalities.229 Property rights are so extensive that
each potential externality is priced in accordance with the consequence to the person who is
affected by it. Of course, who is affected by whose actions is itself determined by the prevailing
system of property rights. It could be, for example, that the body responsible for a wetland, such
as the local authority, is required by law to compensate anyone who suffers from the mosquitos
that breed there (with the compensation calculated in terms of a price equalling
the harm the person suffers from an additional mosquito). In that system, the authority would
ideally also have the right to charge anyone who not only does not suffer from the mosquitos
but instead enjoys the benefits that pollinators bring to their garden (the price being the benefit
to the person from an additional pollinator).230 But if property rights were allocated the other
way round, someone who suffers from mosquitos would have to pay the authority for doing
something about the nuisance.
It is a common observation that well-functioning market systems economise on information.
The institution requires each individual to know only (i) her own mind and her abilities, (ii) the
assets she has a right to, and (iii) the prices of goods and services. She does not need to know
what is on other people’s minds or their abilities, nor what they are doing. But that economy of
information is made possible only by the enormous information burden on the institution itself
– markets have to pool all information and embed them in prices. Prices are interpreted as an
emergent feature of markets (Adam Smith called that the workings of the ‘invisible hand’). You
may possess information of a feature of the world that I do not have, but if I can infer that
information from your actions correctly, then our collective actions will convey private
information through the emerging prices. Box 7.3 showed the implications for resource
allocation when that assumption is relaxed.231
Nevertheless, imagine that the acquisition of information is so cheap that markets are able to
price externalities. Wouldn’t laissez faire then be superior to an institution where regulators
direct some behaviour? Although scholars in the economics of law have argued that laissez faire
can match even well-chosen government regulations on environmental externalities, we show
below that the belief is a stretch. This issue is at the heart of political philosophy, for it has to do
with identifying the tasks citizens would assign to the state in a well-ordered society.232
Mutualism
In a paper that revived the study of externalities in contemporary economics, Meade (1952)
considered a world of beekeepers and fruit growers. Bees pollinate fruit trees while fruit trees
provide nectar and pollen to bees; the two populations raise one another’s productivities.
Ecologists call that kind of population dependence mutualism. Meade derived the structure of
Pigouvian subsidies that would internalise the externalities, and he showed that beekeepers and
fruit growers in the presence of those subsidies would be motivated to increase their stocks to
the point where the mutualism was internalised.
Payment for Ecosytem servuces
As the questions suggest, PES is based on the principle that beneficiaries of ecosystem
services should pay to conserve and restore them. The institution translates externalities into
financial incentives for local actors so that they provide the ecosystem services enjoyed by the
beneficiaries (Chapter 20). PES systems have been much influenced by the fact that landowners
manage their property in ways that are not necessarily (perhaps not even usually) consonant
with the provision of ecosystem services. Modern agricultural practices are a notable example.
Designing PES systems therefore grapple with finding ways in which landowners are to be
compensated for providing the services (Engel, Pagiola, and Wunder, 2008).
Institutions that define PES programmes are often thought of as being ‘markets’, but the
expression should be used with care. They are not markets in the sense that was explored in
Section 7.3. The payments instituted are more like ‘administered prices’ based variously on the
cost of maintaining the ecosystem service, limiting entry to the ecosystem so as not to have
overcrowding, and so on. PES programmes differ in the types and scale of the demand for
ecosystem services, the payment source (i.e. who makes the payment), the type of activity for
which payment is made, the measure used to judge the quality of the service for which payment
is made and of course the size of the payment. The effectiveness of a PES depends crucially on
the programme design.240
As the tropics are rich in biodiversity and many of the world’s poorest countries are in the
tropics (Chapter 14), it has been a natural thought to build PES programmes around poor
communities. A PES system in which the state plays an active role is attractive for wildlife
conservation and habitat preservation. Property rights to grasslands, tropical forests, coastal
wetlands, mangroves and coral reefs are often ambiguous in developing countries. The state
may lay claim to the assets (‘public’ property being the customary euphemism), but if the
terrain is difficult to monitor, inhabitants will continue to reside there and live off its products.
Inhabitants are therefore key players. Without their engagement, the ecosystems could not be
protected. Certain natural ‘wonders’ attract a significant number of tourists on a regular basis
(one may think of the National Parks in North America or the jungles of Borneo as examples).
An obvious thing for the state to do is to tax tourists and use the revenue to pay local
inhabitants for protecting their site from poaching and free-riding. Local inhabitants would then
have an incentive to develop rules and regulations to protect the site. An alternative would be to
hand over the right to charge tourists to the local inhabitants themselves.
There are two aspects of special interest in PES systems in low income countries. One is its
contribution to poverty reduction, the other to biodiversity conservation. Unfortunately, the
findings are mixed on both counts (Zilberman, Lipper and McCarthy, 2008). There are situations
where the system would be bad for poverty reduction and distributive justice. Many of the rural
poor in low income countries enjoy Nature’s services from assets they do not own (Chapter
14). Even though they may be willing to participate in a system of property rights in which
they are required to pay for ecosystem services provided by landowners (Pagiola, Rios, and
Arcenas (2008) reported in their careful study of a silvopastoral project in Nicaragua that they
did), it could be that the economically weaker among them are made to pay a disproportionate
amount. Some may even become worse off than they were previously.243 One could argue that in
those situations the state should pay the resource owner instead, using funds obtained from
general taxation. As Bulte et al. (2008) observe, who should pay depends on the context.
Chapter 21 Options for change, full review