Ex p l o r i n g t h e Po p u l at i o n / Wat e r Re s o u r c e s N ex u s
i n t h e Deve l o p i n g Wor ld
Anthony R. Turton and Jeroen F. Warner
Ant hony R. Turon is head of the African Water I ssues Research Unit at Pretoria University.
Jeroen F. Warner is a researcher at the Flood Hazards Research Centre of Middlesex
University.
Int r oduc t ion
Any discussion on the population/ water resources nexus in the developing world
tends to be clouded by preconceived notions. The very concept of “population growth” is
invariably constructed so as to be linked with value-laden notions like the “population
time-bomb” and “population explosion,” which are, in turn, closely associated with Malthusian catastrophe and social decay. This article explores the population/ water resources
nexus by using empirical examples from Africa in order to isolate some of the strategically important issues that policymakers should recognize. Two distinct areas of Africa
have been selected: first, “Southern Africa,” which for purposes of this article will mean
countries belonging to the Southern African Development Community (SADC); and second, “East Africa,” which for purposes of this article will mean countries geographically
located within the Nile River Basin. 1
The article begins by listing some fundamental points of departure, laying out our
approach to the concepts of scarcity, resources, and legitimacy. These concepts provide
a logical conceptual foundation for the article’s subsequent analysis. This foundation is
followed by an analytical separation between what will be identified as “first-order” and
“second-order” levels of analysis. I t is in this separation that the article offers an alternative perspective on the issues at hand. Special attention will be paid to the use and
usefulness of technologies like Geographic I nformation Systems (GI S) in the context of
this first-order/ second-order framework. The article concludes by answering four key
questions, which are deemed to be important to an understanding of the strategic significance of water resource management in Africa.
Point s of Depart ure: Som e Key Concept s
What immediately follows are fundamental ideas that form the article’s logical foundation. Three concepts of this foundation are of paramount importance: scarcity, resources, and legitimacy.
1. “Water scarcity” seems at first so tantalizingly straightforward as to seduce the
non-specialist reader into a rather superficial understanding of its definition and implications. I sn’t water scarcity simply a shortage of water in time and space? While this definition adequately defines the concept in many cases, the spatial/ temporal dimension of
water scarcity involves subtle but important nuances. Rather, water scarcity should be
defined as a condition in which demographically-induced demand for water exceeds the
prevailing level of local supply (Turton & Ohlsson, 1999). Pressures resulting from absolute population growth and increasing density from urbanization of course raise the number of people per unit area. But a focus on “demand” forces us to take into account how
the notion of scarcity is also economically and culturally constituted. Beyond the three
liters per person per day required for basic human survival, “demand” and even “need”
are not absolute values; they depend on social and consumptive habits that are culture-
52
Po p u l a t i o n a n d W a t e r
bound, differing between countries and within regions.
The availability of water also quite naturally changes with the season. For large parts of
Africa, a drought condition is a totally normal set of circumstances if viewed in terms of
oscillations within the global hydrological cycle. This climatic variability acts as a fundamental driver to many of the African ecosystems in the semi-arid regions, and humans and
other living organisms have adapted to it. The timing and intensity of a flood can make the
same sum total of floodwater a boon or a curse for the cropping season.
Thus, the technical ability with which societies are able to handle their waterresource base is paramount. A perceived condition of water scarcity may in fact exist in the
face of apparent abundance. Current work in Zambia (Turton et al., 2000a), one of the most
well-endowed countries in Southern Africa in terms of water availability, shows how acute
water scarcity can exist even in that country simply because its government lacks the capacity to harness its water in dams and then process and distribute it via an adequate
reticulation system. Heavy water pollution also results in a form of scarcity, sometimes
called “hydrocide” (Lundqvist, 1998). Water quality has a major impact on the functionality
of water: the better its quality, the more social, cultural, economic, and environmental
functions it can perform. So water scarcity is more than just a simple non-availability of
water.
When discussing scarcity, we should also give due cognizance to use and ownership.
Sexton (1992) has differentiated between absolute (i.e. technology-limited) scarcity and
economic scarcity, the latter referring to economic choices that have created winners
and losers. Warner (1992) has noted that the key limit to water availability is redistribution and hence is political rather than technical in nature; and this distinction suggests yet
another distinct form of scarcity—induced . This supports Homer-Dixon’s (1994) resource
capture thesis, which we shall address below.
2. Just as the concept of water “scarcity” is subtler than first meets the eye, we will
also have to come to grips with the nebulous concept of a “resource.” An important point
of departure in this article is that an epistemological and conceptual distinction can be
made between what we will define as a “first-order” and a “second-order” resource. To
our knowledge, Leif Ohlsson (1998; 1999) was the first to systematically analyze resources in this way. I n his analysis, a first-order resource is any natural resource (such as
water, land, or minerals) with which a country can be either well- or poorly-endowed. I n
other words, a first-order resource like water can be either scarce or abundant; and the
degree of scarcity and/ or abundance is relative spatially, temporally, and in terms of
quality. What is stressful in one environment is not a problem in another.
A second-order resource, on the other hand, is a social rather than a natural resource. A social resource refers to a need (acutely perceived by societies, administrative
organizations, and managers responsible for dealing with natural resource scarcities) to
find the appropriate societal tools for dealing with the social consequences of first-order
natural resource scarcities (Ohlsson, 1999, page 161). This conceptual distinction makes
it clear that what is critically important is not so much the availability of the natural
resource itself but rather how society adapts to changes in that supply—either by way of
(a) long-term increases in water scarcity as a result of population growth and/ or climate
change, or (b) short-term water abundance in the form of floods.
Recent articles using this distinction depict water management as a series of oscillations between first and second-order resources over time, much like the turning of a
screw (Ohlsson & Turton, 1999; Ohlsson & Lundqvist, 2000). Priorities change from supply-sided management (mobilizing more water) through demand-sided management (doing
better things with available water), ultimately to adaptive management (adapting to
absolute scarcity). Couched differently, Ohlsson’s (1998; 1999) second-order resource is
Po p u l a t i o n / W a t e r N e x u s
53
another way of looking at Thomas Homer-Dixon’s (1995; 1996) concept of “ingenuity.” But
the importance of this conceptual difference is that it allows the analyst and policymaker to
effectively develop coping strategies to deal with the bottlenecks inherent in water management globally. This has particular relevance for an understanding of the problems confronting developing countries.
This conceptual distinction makes it possible to develop a whole range of unique
Fi g u r e 1 . Re s o u r c e M at r i x
Quant it at ive Aspect of
t he Resource
Type of Resource
Second-Order
First Order
(Social Resources)
(Water Resources)
Relat ive
Abundance
Posit ion 1
Posit ion 2
Relat ive
Scarcit y
Posit ion 3
Posit ion 4
concepts by means of a matrix showing different levels of first- and second-order resources within any given social entity. This is illustrated in Figure 1.
Four combinations of first- and second-order resource are possible. For purposes of
this article, only the last three of these combinations (those entailing at least one relative
scarcity) are relevant:
• St ruct urally-I nduced Relat ive Wat er Scarcit y ( SI RWS) is a combination
that consists of a relatively high level of first-order resource availability (Position 1) in
conjunction with a relatively low level of second-order resource availability (Position
4). Water scarcity in these situations is probable as a result of the inability to mobilize sufficient social resources to effectively manage the problem. SI RWS countries
are relatively well-endowed with water, but lack institutional capacity and have other
problems that render them unable to mobilize that water (via dams and related
hydraulic infrastructure) and reticulate it to the end-user. A logical outcome of this
condition would be low economic activity, poor public health, and a general low level
of infrastructural development. This condition is clearly unfavorable, and could result
in a Malthusian catastrophe if combined with high population growth. But creative
54
Po p u l a t i o n a n d W a t e r
and responsible decision-making can still save the day provided that the alarm bells are
heeded in time. I t is these societies that offer examples of the debilitating effects of
Homer-Dixon’s (1995; 1996; 2000) “ingenuity gap.” Examples include Angola, Congo
(DRC), Mozambique, and Zambia.
• St ruct urally-I nduced Relat ive Wat er Abundance ( SI RWA) refers to a combination that consists of a relatively low level of first-order resource availability (Position 3) with a relatively high level of second-order resource availability (Position 2). I n
other words, water abundance is made possible in a relative sense as a result of the
ability to mobilize sufficient social resources to effectively manage the problem. SI RWA
countries are relatively poorly endowed with water resources, but use their relative
abundance of social resources to develop a set of management solutions that are
effective and legitimate in the eyes of the population and therefore sustainable over
time. A logical outcome of this condition would be sustained economic growth, good
public health, and a high level of infrastructural development even in the face of
endemic water scarcity. This condition resembles the Cornucopian argument that is
often presented as an alternative to Malthusian collapse. I ndeed there are rich examples of the positive impact of Homer-Dixon’s (1996; 2000) concept of ingenuity to
be found in an analysis of the water sector in many countries. Arguably the best
example is I srael, but South Africa occupies a close second in this category.
• Wat er Povert y ( WP) refers to a combination that consists of a relatively low level
of first-order resource availability (Position 3) with a relatively low level of secondorder resource availability (Position 4). WP countries cannot manage the debilitating
effects of water scarcity because of their lack of social resources, unleashing a spiral
of underdevelopment that results in a gradual decline in almost all developmental
indicators. A logical outcome of this condition would be long-term economic stagnation, deteriorating public health, a low level of infrastructural development, and a
high probability of social instability and political decay as the black hole caused by a
combination of expanding population and a declining resource-base takes hold. I n
short, this is an example of the classic Malthusian collapse. Clearly this condition is
one to be avoided.
3. Finally, “legitimacy” (which can loosely be defined as the popular support by the
broad population for any given decision by government) is an important concept for
effective water management. For Water Demand Management (WDM) policies to be
effectively implemented, a high level of legitimacy is required of the functional agency
responsible for water-resource management ( Turton, 2000a, page 144); yet that
government’s craving for legitimacy easily leads to policies that have the opposite effect.
I n many political systems, intersectoral allocation of water (Turton & Ohlsson, 1999;
Turton, 1999; Allan, 2000, page 184) is typically considered only as a last resort because
it is so politically and socially risky that politicians generally favor softer (but also less
effective) options instead.
When river basins reach closure and all available first-order resources have been
allocated, one of the most important forms of management strategy—after all other
supply-sided options such as I nter-Basin Transfers (I BTs) and desalination of water have
been exhausted—is the allocation of water away from high-consumption but low-yield
activities (as typically found in the agricultural sector) to lower-consumption but higheryield activities (typically found in the industrial and domestic sectors) (Falkenmark &
Lundqvist, 1995). There are a number of unintended consequences of this, such as those
arising from new economic dependencies and the restructuring of society away from an
agricultural base to an industrial base. Whether this will actually happen depends in part on
the second-order capacities and structures for change that exist in society. But as the public
Po p u l a t i o n / W a t e r N e x u s
55
sector tends to be lead actor and regulator as well as often the formal owner of water
resources, a successful adaptation to first-order stress also depends on the relationship
between the state and society. A power relationship is legitimate when the relationship can
be justified in terms of people’s beliefs—when there is congruence between power and
beliefs, values and expectations (Weber, 1947).
I f people already believe in the need for an adaptive response to water stress, and if
the government’s legitimacy base is strong, a society will be more responsive to regulatory measures aimed at bringing about this adaptation. I f these values are not strongly
developed, a government perceived as legitimate may well still have the political capital
to guide society to a new mindset. However, if a ruling government perceives that it lacks
legitimacy, it may not be willing to take the political risk of implementing unpopular
policies, even when the society faces an uncontrolled and ultimately unsustainable spiral
of water consumption. The state may be tempted to pursue wasteful but popular water
projects instead.
The world is filled with examples of ill-considered water projects that have been used
to buy political support, otherwise known as patronage. Specific examples range from to
pork-barrel projects in the United States, the Pongola-Poort Dam in South Africa, and
many instances in I ndia where unsustainable water projects cannot be changed because
they are supporting too many jobs and therefore potential voters. This situation is found
in several postcolonial states, which started large, unsustainable projects to kick-start
economic development. When these aspirations come to nothing, the government starts
losing the political capital needed to make social adjustments to water policy that address an eroded and unsustainable resources base. As Ohlsson (1999, page 10) notes,
the first victim of people’s frustrated developmental expectations is state legitimacy. I ncidentally, this is not limited to the developing world. The so-called “pork-barrel projects”
in the United States that Reisner (1993) so eloquently describes illustrate patronage in a
sophisticated democratic setting.
Finally, a situation is conceivable in which society may have a latent willingness and
ability to adapt, but systemic legitimacy (of the political system itself ) is sorely lacking. I n
apartheid-era South Africa, for example, all reform was hampered by the systemic illegitimacy of the system itself, resulting ultimately in a collapse and radical restructuring of
the overall political process. A decision-making entity perceived as illegitimate will not
receive the necessary popular support, and the population at large will undermine such
government policies as a form of civil disobedience. Notably, in the developing world, we
find examples of governments and implementing agencies with a low level of accountability and consequently a low level of systemic legitimacy. I nstead of initiating reflexive
change, these governments and agencies tend either to ignore the water crisis or to
deflect it by further squeezing their natural resource bases, often in the form of intensified production (otherwise known as “water mining”).
Under such stress, the process that has been called “resource capture” (HomerDixon, 1994; Homer-Dixon & Percival, 1996) is especially prone to manifest itself. This
occurs when powerful groups in society systematically shift (first-order) resource allocation in their favor over time, usually to the long-term detriment of the group from which
the resource base is being captured. Since these powerful groups must gain control over
the resource allocation mechanisms in order to gain such unequal access, structural
scarcity (a highly specific form of water scarcity) ultimately results. A good example of
structural scarcity is apartheid-era South Africa, whose “hydraulic mission” effectively
mobilized water in order to distribute patronage to the white minority, thereby retaining
the support of the white farmers who owned most of the land at that time (Abrams,
1996; Turton, 2000a, page 142).
I n international river basins, countries may also try to shift the burden of resource
56
Po p u l a t i o n a n d W a t e r
closure (that condition when all of the resource-base has been allocated) to other riparians.
Upstream riparians may capture the resource before it reaches the downstream countries, while downstream countries may strengthen their claim to a river’s water by leveraging non-water power threats (Warner, 1993). The importance of water is blown out of
proportion under these circumstances, and hydrological information may even be classified because of it. This process propels water management into a national security issue
in which the resource becomes non-negotiable, forestalling an equitable agreement on
its sharing. This “securitization” of water, often an unsatisfactory state of affairs, leads to
zero-sum hydropolitical dynamics.
One possible way of accomplishing a desirable de-securitization of the water issue is
(a) to develop uncontested data with which to build confidence between riparian states
or water users, and (b) to institutionalize the conflict potential that arises under conditions of scarcity. According to Haas (1993), “epistemic communities” may converge around
a body of accepted scientific procedure and thus facilitate the creation of a legitimate
base for negotiation. The creation of water regimes can therefore be seen as a manifestation of second-order resources within any given regional security complex. Processes of
the securitization of data (Warner, 2000) can still obstruct the dissemination and exchange of reliable hydrological information within the emerging regime, however, and act
as a mitigating factor. I n I srael, for example, hydrological data are classified as secret and
is thus not available to the public or other interested parties. This article will later address
the issue of whether GI S can enhance openness and data exchange, thereby facilitating
confidence building in water-sharing arrangements.
T h e Po p u l at i o n /Wat e r Re s o u r c e N ex u s i n A f r i c a
First -Order Type of Analysis
I f water resources are relatively finite within any given country, then a doubling of
that country’s population will cut in half the volume of water available per capita. This
calculation is seductively simple, so let us don the eyeglasses of first-order analysis and
look at some African countries. Table 1 shows the population data for Southern African
and East African countries in Columns 2-6. The population growth over that time period
(39 years) is shown as a percentage in Column 7 as calculated by the FAO (2000) database. I n general terms, this table gives an indication of how the baseline population,
which was arbitrarily taken as being 1961, had grown by 2000. Column 8 shows the
water availability expressed in cubic meters per capita in 1998. The World Bank Atlas
(2000, page 30) defines water availability per capita as the total renewable water resources of a given country (including river flows) divided by the population and expressed
in cubic meters.
Two assumptions (both of which are strictly of a first-order nature) can be made for
purposes of analyzing this data. The first assumption is that a three percent growth in
population over a 39-year period is high. All of the countries that have a population
growth in excess of this have been listed in Table 1 in bold, and their corresponding
figures in Column 7 have been highlighted. The second assumption is that, in terms of
availability of fresh water, anything above 10,000 cubic meters of water per person per
annum is high. Here, too, the countries concerned have been listed in bold, and their
corresponding figures in Column 8 have been highlighted. Clearly, these assumptions are
rather arbitrary and can be challenged. But it should also be emphasized that (a) these
assumptions are relative and not absolute, and (b) they establish a clear split in the data
in order that basic trends can be detected (see the following discussion on GI S). As such,
they act as filters enabling raw data to be analyzed in some meaningful ways (see Box 1 for
a full explanation of the methodology used in this circle).
Po p u l a t i o n / W a t e r N e x u s
57
Ta b l e 1 . Po p u l at i o n Gr ow t h (M i l l i o n s ) a n d
Wat e r Av a i l a b i l i t y Dat a
Sout hern Africa ( SADC Mem ber St at es)
1961
1970
1980
1990
2000
Growth
19612000 (%)
A va ila ble
m 3/cap
1998
A ngola
4 .8
5 .5
7.0
9 .2
1 3 .1
2 .6 8 %
1 5 ,7 8 3
Botswa na
.4 9
.6 3
.9 0
1 .2
1 .5
3 .1 8 %
9 ,4 1 3
1 5 .7
2 0 .2
27.0
37.3
5 0 .9
3 .1 1 %
2 1 ,1 3 4
Lesotho
.8 8
1 .0 6
1 .3
1 .7
2 .0
2 .1 3 %
2 ,5 2 7
Ma la wi
3 .6
4 .5
6 .1
9 .3
1 1 .3
3 .2 0 %
1 ,7 7 5
Mauritius
.6 7
.8 2
.9 6
1 .0
1 .1
1 .3 1 %
1 ,8 9 7
Moz a mbique
7.6
9 .3
1 2 .0
1 4 .1
1 8 .3
2 .2 5 %
1 2 ,7 4 6
Na mibia
.6 3
.7 9
1 .0
1 .3
1 .7
2 .7 3 %
27,373
Seychelles
.4 3
.5 3
.6 3
.7 0
.8 0
1 .4 9 %
n/a
South A frica
17.8
2 2 .0
27.5
3 4 .0
4 3 .3
2 .3 5 %
1 ,2 0 8
Swaziland
.3 3
.4 1
.5 6
.7 5
.9 2
2 .7 9 %
4 ,5 5 2
T a nz a nia
1 0 .4
1 3 .6
1 8 .5
2 5 .4
3 5 .1
3 .2 2 %
2 ,7 7 0
Za mbia
3 .2
4 .1
5 .7
7.2
1 0 .4
3 .1 6 %
1 2 ,0 0 1
Zimba bwe
3 .9
5 .2
7.1
9 .8
1 2 .6
3 .2 6 %
1 ,7 1 1
Country
Congo (DR)
East Af rica ( Nile Basin St at es)
Country
1961
1970
1980
1990
2000
Growth
since 1961
A va ila ble
m 3/cap
1998
2 .9
3 .5
4 .1
5 .4
6 .3
2 .1 6 %
561
Eg y p t
2 8 .5
3 5 .2
4 3 .7
5 6 .3
67.8
2 .2 9 %
949
Eritrea
n/a
n/a
n/a
n/a
3 .6 5
2 .1 9 % * *
2 ,2 6 9
Ethiopia
2 4 .7
3 0 .6
3 8 .7
5 0 .9
6 2 .9
2 .4 3 %
1 ,7 9 5
Ke nya
8 .5
1 1 .4
1 6 .6
2 3 .5
3 0 .6
3.51%
1 ,0 3 1
Rwanda
2 .8
3 .7
5 .1
6 .9
7.6
2 .2 1 %
798
Soma lia
2 .8
3 .6
5 .8
7.7
8 .7
3 .0 5 %
1 ,7 3 0
Sudan
1 1 .3
1 3 .8
1 8 .6
2 4 .0
3 1 .1
2 .6 5 %
5 ,4 3 3
T a nz a nia
1 0 .4
1 3 .6
1 8 .5
2 5 .4
3 5 .1
3.22%
2 ,7 7 0
Uga nda
6 .8
9 .8
1 3 .1
1 6 .4
2 3 .3
3.11%
3 ,1 5 8
Burundi
Sources of Dat a:
Population growth data (Columns 2 - 6) - FAO (2000).
Population growth since 1961 (Column 7) - FAO (2000).
Water availability per capita m 3 in 1998 (Column 8) - World Bank Atlas (2000, page 34-35).
* * Eritrea Data calculated from 1993 - FAO (2000).
58
Po p u l a t i o n a n d W a t e r
From this rather crude assessment, an interesting picture starts to emerge. All of the
countries that have a relatively low population-growth (i.e. less than a three percent increase in 39 years) in conjunction with a relatively high availability of freshwater are found
in Southern Africa and include Angola, Mozambique, and Namibia. Conversely, countries
that have a relatively high population-growth rate in conjunction with relatively low wateravailability include Botswana, Malawi, Tanzania, Zimbabwe, Kenya, Somalia, and Uganda.
Two countries—the Democratic Republic of the Congo (DRC) and Zambia—stand out alone
in terms of this assessment, displaying a relatively high population growth rate in conjunction with a relatively high water-availability.
The analysis also suggests that the countries in the first group (low population growth
and high water-availability—Angola, Mozambique, and Namibia—have population and water-resource fundamentals that ought to predispose them to a degree of prosperity. But
this is not the case. While Angola is richly blessed with a wide range of first-order resources, it remains embroiled in a debilitating civil war; if anything, its resources (particularly oil and diamonds) only fuel the conflict. Mozambique has a seemingly high volume of
water resources, yet it too has suffered from a quarter-century of civil war and is in fact a
downstream riparian on almost all of the river basins passing through it, making it vulnerable to the whims of upstream states. Namibia is relatively prosperous and politically
stable, yet it has a small population and therefore a small tax base—a debilitating factor.
Namibia’s physical size is massive and its population far removed from water resources,
meaning that any infrastructural projects actually have a low number of taxpayers per
kilometer of pipeline.
The second group of countries (high population growth and low water-availability)
suggests future Malthusian catastrophes in each country except for Botswana. Botswana
is actually one of the most politically stable countries in Africa; it has a functioning multiparty democracy, and its high population growth rate is off a low original population base,
so there are not in this case the normal problems related to a rapidly growing population.
Significantly, Botswana is also adopting progressive water policy options that include the
preference for food security rather than national self-sufficiency. Malawi, Tanzania, Zimbabwe, Kenya, Somalia, and Uganda have all had histories of political instability and
economic stagnation, although this is changing for Tanzania and Uganda. First-order
types of analysis in these cases are clearly superficial and can be downright dangerous:
it is from this type of analysis that the so-called “water wars” literature derives its empirical basis. 2
Kaplan’s (1994) presentation of sub-Saharan Africa as an anarchic, conflict-ridden
basket-case has raised suspicion (Ó Tuathail et al, 1998) that such accounts fit certain
North American national security agendas. Whatever the agenda, the image of an imploding and chaotic Africa undermines investor confidence in the continent and marginalizes
it in policy debates as a lost cause. The Kaplan thesis assumes that water resources are
finite, which they are not. Those such as Gleick (1993) who proclaim a “water crisis” only
focus on that fraction of precipitation that ends up in rivers, lakes, and aquifers (what has
come to be known as Falkenmark’s “blue water”). Even in this case, baseline data often
unreliable. The remainder either evaporates to the atmosphere, is taken up by vegetation, or percolates into the soil, where it remains as soil moisture and lies unaccounted
for in the “water crisis” vernacular. This latter amount (which has become known as
Falkenmark’s “green water”) is so abundantly available in the temperate zones that these
areas can export huge quantities of “encapsulated water” in the form of grain to semiarid zones that are structurally deficient in soil moisture. On the basis of this principle,
Allan (1996; 2000) has shown that trade in virtual water —the water embedded in cereals—is a viable alternative provided that sufficient foreign currency can be generated to
pay for such exports (see Box 2). Allan notes that as much water flows into the Middle
Po p u l a t i o n / W a t e r N e x u s
59
East North Africa (MENA) region as subsidized grain in the form of virtual water as flows
down the Nile annually. I t is this trade in virtual water that has helped prevent the
confidently-predicted water wars (Starr, 1991) from erupting (Turton, 2000b; Allan, 2000).
I n addition to ignoring these international economic processes, the water-in-crisis
thesis misunderstands the nature of “resources” that are often interpreted in environmentally deterministic ways long since abandoned in geography (Bradnock & Saunders,
2000). Such an analysis simply ignores the capacity of states to develop coherent and
Box 1 . M e t h o d o l o g y
T
he methodology that has been used in this article is based on four assumptions, each of which has been arbitrarily defined. The purpose of these assumptions is to act as a type of filter through which raw data can be processed in order
to arrive at a conclusion that can assist with the development of a set of core
hypotheses. These hypotheses can then be used in other case studies, in order to
test their validity, but also in order to refine the underlying concepts and thereby
develop new knowledge. This is necessary because the notion of a second-order
resource is relatively new and consequently in need of conceptual refinement.
These four assumptions are as follows:
• The first assumption is that a 3 percent growth in population over a 39-year
period is High, with a growth below this level being considered as Low. This
is an arbitrary selection in order to give us a starting point in our analysis.
Because of the contested nature of population figures in developing countries, the data from the FAO (2000) is being taken as the legitimate source.
• The second assumption is that in terms of the availability of fresh water within a
given country, anything above 10,000 m3/ cap/ yr -1 is High, with anything less
than this value being considered as Low. The data used are derived from the
World Bank (2000:34-35) because such data are highly contested in the
developing world, and the criterion for the High/ Low split is arbitrarily defined
in order to give us a starting point for the analysis.
• The third assumption is that a GNP/ cap when adjusted to Purchasing Power
Parity (PPP) for any given country as defined by the World Bank (2000:42-43)
is considered to be High if above a value of US$5,300. Conversely a value
below US$5,299 is considered to be Low.
• The fourth assumption is that with respect to the percentage of a national
population that has access to relatively safe drinking water as defined by the
World Bank (1999), a value greater than 65 percent is considered to be High,
with a value below 64 percent being Low.
I t must be noted that these assumptions are not ironclad. I n reality data is
highly contested in the developing world, and these will be no exception, which
means that the debate normally degenerates into one about the unreliability of the
figures being used. This is a sterile debate; so in order to make some headway in
our quest for the development of new knowledge, these four assumptions have
been made. They should not be seen as being concrete in any way, but when used
in combination form a valid methodology on which the rationale of this article has
been based. This methodology enables us to steer a reasonably safe course
through the minefield of unreliable data that are a characteristic of the developing
world, and it enables us to compare countries in a meaningful way.
60
Po p u l a t i o n a n d W a t e r
Ta b l e 2 . Co m p a r i s o n o f Fi r s t a n d
Se c o n d -Or d e r Re s o u r c e s
Sout hern Africa ( SADC Mem ber St at es)
Country
First -Order I ndicat ors
P opula tion
Growth
(since 1961)
a
(b)
Second-Order I ndicat ors
Wa te r
A va ila bility
m 3/cap/yr-1
1998
a
(b)
GNP /ca p
U S$ Purchasing
Power Parity
1998
a
(b)
A cce ss of
P opula tion to
Safe Water %
a
(b)
A ng o l a
2.68% (low)
15,783 (high)
999 (low)
32% (low)
Botswana
3.18% (high)
9,413 (low)
5,769 (high)
70% (high)
C ongo (DR)
3.11% (high)
21,134 (high)
733 (low)
27% (low)
Lesotho
2.13% (low)
2,527 (low)
2,194 (low)
52% (low)
Malawi
3.20% (high)
1,775 (low)
551 (low)
45% (low)
Mauritius
1.31% (low)
1,897 (low)
8.236 (high)
98% (high)
Mozam bique
2.25% (low)
12,746 (high)
740 (low)
32% (low)
Nam ibia
2.73% (low)
27,373 (high)
5,280 (low)
57% (low)
Seychelles
1.49% (low)
n/a
10,185 (high)
97% (high)
South A frica
2.35% (low)
1,208 (low)
8,296 (high)
70% (high)
Swaziland
2.79% (low)
4,552 (low)
4,195 (low)
43% (low)
Tanzania
3.22% (high)
2,770 (low)
483 (low)
49% (low)
Zam bia
3.16% (high)
12,001 (high)
678 (low)
43% (low)
Zim babwe
3.26% (high)
1,711 (low)
2,489 (low)
77% (high)
East Africa ( Nile Basin Riparian St at es)
Country
First -Order I ndicat ors
P opula tion
Growth
(since 1961)
a
(b)
Second-Order I ndicat ors
Wa te r
A va ila bility
m 3/cap/yr-1
1998
a
(b)
GNP /ca p
U S$ Purchasing
Power Parity
1998
a
(b)
A cce ss of
P opula tion to
Safe Water %
a
(b)
Burundi
2.16% (low)
561 (low)
561 (low)
52%
Eg y p t
2.29% (low)
949 (low)
3,146 (low)
64%
Eritrea
n/a
2,269 (low)
948 (low)
7%
Ethiopia
2.43% (low)
1,795 (low)
566 (low)
27%
Kenya
3.51% (high)
1,031 (low)
964 (low)
53%
Rwanda
2.21% (low)
798 (low)
n/a
n/a
Som alia
3.05% (high)
1,730 (low)
n/a
37%
Sudan
2.65% (low)
5,433 (low)
1,240 (low)
50%
Tanzania
3.22% (high)
2,770 (low)
483 (low)
49%
U ganda
3.11% (high)
3,158 (low)
1,072 (low)
34%
Po p u l a t i o n / W a t e r N e x u s
61
Sources of dat a for Table 2:
Populat ion grow t h since 1961 (Column 2) - Column 7 of Table 1.
High/ Low Population growth split (Column 2) - Arbitrarily defined as > 3.0% is high,
< 2.9% is low.
Wat er availabilit y m 3/ cap/ yr -1 1998 (Column 3) - World Bank Atlas (2000, pages 3435) and Column 8 of Table 1.
High/ Low water availability (Column 3) - Arbitrarily defined as > 10,000 m 3/ cap/ yr -1 1998
is high, < 9,999 m 3/ cap/ yr -1 1998 is low.
GNP/ cap 1998 (Column 4) - World Bank (2000, pages 42-43)
High/ Low GDP/ cap split (Column 4) - Arbitrarily defined as > $5,300 is high, < $5,299 is
low.
Access of Populat ion to Safe Water (Column 5) - World Bank (1999).
High/ Low Access of Population split (Column 5) - Arbitrarily defined as > 65% high,
< 64% is low
sustainable policy choices with which to manage the problem of water scarcity. I t is this
type of capability that fits into the category of “second-order resources,” which can loosely
be defined as the social resources needed to manage changes in the level of first-order
natural resource availability—otherwise known as social adaptive capacity—over time.
Second-Order Type of Analysis
When it comes to second-order analyses, we are confronted with a basic problem.
How do we identify and measure social adaptive capacity? How do we know when it
exists and when it is absent? These questions are currently the subject of a research
project at the African Water I ssues Research Unit (AWI RU) (Turton et al., 2000a; Turton,
2002; Turton & Kgathi, 2002). Their answers require a set of indicators of second-order
resource presence (or absence). Again, one needs to make certain assumptions in order
to gain insight. For the purposes of this article, two key indicators will be used:
• Let us assume that the existence of second-order resources will result in a higher
degree of economic prosperity than the absence of those resources, in line with
Homer-Dixon’s (1995; 1996; 2000) ingenuity thesis. I f this is true, then the adjusted
GNP per capita at Purchasing Power Parity (PPP) as presented by the World Bank
(2000, pages 42-43) can be used as an indicator.
• The percentage of a given national population that has access to reasonably safe
drinking water is an indicator of a government’s capacity to provide basic services.
World Bank (1999) data on these percentages will be used as an indicator.
Table 2 presents these indicators in the following sequence. Column 1 of the table
names the country concerned. First-order indicators are presented in Columns 2 and 3.
Column 2a shows the population growth rate for that country as shown in Column 7 of
Table 1. This provides an indicator of the country’s population dynamics over the last 39
years, which is shown as a High/ Low split in Column 2b. (See the first assumption in the
previous section for a discussion of the criterion for this split.) Column 3a presents the
availability of first-order water resources per capita expressed as cubic meters per annum
as shown in Column 8 of Table 1. Column 3b shows this data as a High/ Low split (using
the second assumption that is based on the criterion discussed in the previous section).
This provides a crude but useful indicator of first-order water resource availability assuming that the country can develop those resources.
62
Po p u l a t i o n a n d W a t e r
Ta b l e 3 . Cl a s s i f i c a t i o n o f Va r i o u s A f r i c a n St a t e s i n
t er m s o f Pr o p o s e d Ty p o l o g y
Southe rn A frica
East A frica
First-Orde r
P roble ms
Se cond-Orde r
P roble ms
More Comple x
P roble ms
SIRWA
SIRWS
WP
Botswana
Mauritius
South A frica
A ng o l a
Congo (DRC)
Mozam bique
Nam ibia
Zam bia
Lesotho
Malawi
Swaziland
Tanzania
Burundi
Eg y p t
Eritrea
Ethiopia
Kenya
Sudan
Tanzania
U ganda
Second-order indicators are presented in Columns 4 and 5 of the table. Column 4a
shows the GNP per capita as US dollars adjusted in terms of Purchasing Power Parity
(PPP). Column 4b presents this data as a High/ Low split, with the criterion arbitrarily
defined as > $5,300 being High and < $5,299 being Low (our third assumption). While this
is an unsophisticated way of processing the data, it serves as a filter that shows an
ultimately useful relative tendency. Column 5a shows the percentage of a given national
population that has access to relatively safe water. Column 5b presents this data as a
High/ Low split, with the criterion arbitrarily defined as > 65% being high and < 64% being
Low (our fourth assumption). This is also crude, but serves the same purpose of filtering
out a general tendency. The combination of these indicators (when subjected to the
High/ Low filtering process) can then form the foundation of a hypothesis that can later be
empirically tested. (Again, see Box 1 for a full explanation of this article’s methodology.)
By concentrating exclusively on Columns 3-5, an assessment can be made using the
following logic. Suppose one (mistakenly) assumed that first-order resource abundance
(an independent variable) naturally predisposes a country to economic prosperity (a
dependent variable). One would then anticipate finding a rough correlation in terms of
High/ Low splits between Columns 3 and 4. A cursory glance at Table 2 will show that this
is not the case; so one can conclude that first-order resource abundance on its own is an
insufficient condition to guarantee economic prosperity—suggesting that some form of
interceding variable is at work. I f this interceding variable is expressed in terms of a
second-order resource, then a comparison of Columns 4 and 5 reveals that in all cases
except one (Zimbabwe) the existence of such resources as reflected by a higher GNP per
capita determines the capacity of the government to deliver basic services like the provision of clean water.
Here the logic of Homer-Dixon’s (1995; 1996; 2000) ingenuity thesis is relevant. The
presence of a higher level of second-order resource translates into a higher level of
economic activity, which in turn impacts on the ability of the state to deliver basic services. Botswana offers a revealing insight in this sense. A country with a relatively small
population size but a high population growth rate, it faces severe constraints in terms of
Po p u l a t i o n / W a t e r N e x u s
63
low water-availability, yet still maintains a high level of service delivery. A similar trend is
evident in Mauritius and South Africa, where high levels of service delivery are possible
despite severe first-order water constraints. Namibia is also revealing. A small population in absolute terms usually impacts on the availability of water by showing a high
potential for development. I n Namibia, however, a low level of economic activity (coupled
with a small tax base) acts as a severe constraint that is reflected in the country’s low
level of service delivery. Namibia and Botswana also both lack permanently flowing rivers
within their borders, leaving their hinterlands dry and consequently difficult to develop.
Both countries also have a relatively small population and consequently a small tax base.
(The fact that the GNP/ capita indicator is split differently for these two countries is probably irrelevant, given the crudeness of the criterion used and the arbitrary selection of the
threshold at $5,300—see Table 2.)
Applying this methodology to Table 2 yields a neat differentiation of cases consistent
with the key concepts presented at the start of this article. Particular emphasis is placed
on the three conditions: SI RWA, SI RWS, and WP. This typology is presented in Table 3.
Table 3 shows that the typology manifest in the concepts of SI RWA, SI RWS, and WP
can be applied to all cases for which data are available—with only one exception. Zimbabwe presents an anomalous situation that does not fit neatly into this framework: it has
a combination of low levels of both first- and second-order resources, but a high level of
service delivery. While Zimbabwe’s current political leadership has had a negative impact
on the economy, creating an acute shortage of second-order resources, the country’s
high levels of service delivery are manifestations of early Mugabe-era achievements.
Zimbabwe still has a high potential for development, provided that the negative ramifications of its poor political leadership can be resolved.
The matrix’s analysis of Southern Africa yields results that correspond well with each
country’s first- and second-order resource rating. The three SI RWA cases in Southern
Africa are known to be the most prosperous countries in the region. (Should data have
been available for Seychelles, then this country would probably also fall into this category.) For these countries, water-related problems are primarily of a first-order nature—
namely, the continued search for and mobilization of alternative sources of water supply.
The relative economic prosperity of these countries affords them a wide range of options,
covering supply-sided solutions (i.e., development of ever-more-distant water resources
via I BTs and desalination where appropriate), management of demand, and the importation of virtual water in an attempt to balance national water budgets. I ndeed, these
countries are enacting all three strategies (Turton et al., 2000b).
The five SI RWS cases are all countries that ostensibly have an abundance of water
but that lack the institutional, financial, or intellectual capital to translate this into economic growth and development. As such, the type of problems facing these countries are
primarily of a second-order nature. Angola and the Democratic Republic of the Congo
(DRC) are politically unstable because of seemingly endless civil wars. Mozambique has
turned its back on civil war and is seemingly on the road to economic recovery; its
institutional capacity, however, is extremely weak, and its high debt burden continues to
hamper this recovery. The major floods that took place in Mozambique in early 2000 set
the country back significantly economically (Christie & Hanlon, 2001) and also illustrated
the government’s inability to respond to crisis. Namibia is politically stable, but it has
become embroiled in the wars in Angola and the DRC and is starting to hemorrhage
precious financial resources that could be used on institutional development instead.
Namibia also presents an interesting case in the sense that its first-order type of indicators shows the country to be relatively well-endowed with water. However, this water can
only be found on the northern and southern borders of the country and is also difficult to
mobilize. Namibia’s low population levels also create a false impression by presenting a
64
Po p u l a t i o n a n d W a t e r
Box 2 . Tur t o n /Oh l s s o n Gr i d
Water Secure
26
24
Congo DR
22
20
Namibia
18
Angola
16
14
Mozambique
Zambia
12
Adaptively Secure
10
9
8
7
Adaptively I nsecure
6
5
4
3
2
1
Botswana
Tanzania
Malawi
Eritrea
Sudan
Swaziland
Uganda
Lesotho
Ethiopia
Zimbabwe
Kenya
Egypt
Burundi
1
2
3
4
Water I nsecure
First -Order Resource Availabilit y Expressed as
Freshw at er Availabilit y in 1998 m 3 / cap/ yr -1 X 10
28
5
Mauritius
South Africa
6
7
8
9
10
Data Source: World Bank Atlas (2000).
The linkage between water availability and development was drawn directly from the
pioneering work by Malin Falkenmark, who sought to develop a scale with which to measure what she called “water stress.” Her work makes a direct linear relationship between
water availability and the capacity for economic development within a given political economy.
Stated simplistically, Falkenmark (1986) hypothesized that water scarcity presents a
rigid barrier to economic and social development. She sought to measure this by doing an
analysis of various countries in which she found the following: I raq uses 4,400 m3/ cap/
-1
-1
-1
yr ; Pakistan uses 2,200 m 3/ cap/ yr ; Syria uses 1,300m 3/ cap/ yr ; Egypt uses 1,200
-1
-1
-1
m 3/ cap/ yr ; I ndia uses 800 m 3/ cap/ yr and I srael uses 500 m 3/ cap/ yr (Falkenmark,
1986, page 197). By taking I srael as a baseline case, Falkenmark concluded that a
-1
-1
realistic level for a developing state is m 3/ cap/ yr , as this would allow 100 m 3/ cap/ yr
-1
for domestic and industrial use, leaving the remaining m 3/ cap/ yr (80 percent of the
total) for irrigation. I n the quest to develop a scale based on standard units of measurement, Falkenmark then converted this baseline volume (500m 3) to 2,000 people per
“flow unit” of water (one million m 3 of water per year). This lead her to conclude that
more than 2,000 people per “ flow unit” would preclude a region or country from
having sust ainable economic or social development . While not direct ly st at ed by
Falkenmark, this notion implied that water scarcity “beyond the barrier” would result in
social decay and possibly political instability. The notion also contributed to the “water-
(Continued on page 66)
Po p u l a t i o n / W a t e r N e x u s
65
wars” literature, in which this linear relationship was assumed to mean that countries will go
to war over water scarcity at some time in the future.
The analysis seemed intuitively useful at the time, but subsequent research has shown
that countries with a large volume of water available to them do not necessarily develop economically. Conversely, countries with a limited water supply (such as I srael, Botswana, and
South Africa) are capable of economic development close to or even beyond the hypothetical
barrier. I n contrast, the concept of the second order resource has proven pivotal in translating
water availability into economic development. When second-order issues are considered, how
a social entity deals with a scarcity rather than the scarcity itself becomes the critical issue.
Thus, emphasis shifts away from the analysis of water availability (as a first order resource) to
social adaptive capacity (Ohlsson, 1999; Ohlsson & Turton, 1999) or “ingenuity” (Homer-Dixon,
1995; 2000). Water scarcity itself is seen as a relative thing, with a variety of forms in existence
(Turton, 2002).
But what actually is a second order resource? Turton & Ohlsson (1999) developed a grid
showing different combinations of first- and second-order resources, and from this they generated some new conceptual definitions. This was refined by Allan (2000, page 324) and used to
explain the Middle East North Africa (MENA) water situation. There is a strong need to refine
the notion of second order resources further. Five key indicators of second order resources are
currently being developed by Turton (2002):
• GNP/ Capita adjusted to Purchasing Power Parity. This is a crude indicator of potential
for institutional development in a given country and allows a rough comparison to be made
between countries (as shown in the attached grid).
• The ability to generate data is a direct indicator of technical ingenuity. Data are needed
to support decision-making, and contested data are often at the root of hydropolitical
tension between riparian states in shared river basins.
• The ability of a given country to generate coping strategies with which to manage
water scarcity is an indicator of both social and technical ingenuity. I n this regard, certain
issues (such as a policy change from national self-sufficiency in foodstuffs to food security)
are a key indicator. Food security requires foreign currency with which to purchase “virtual
water” embedded in cereals and thereby balance out the national water budget in a “politically silent” way. I n other words, water scarcity is actually a local issue rather than a global
one if one re-defines the problem. Underlying this is what has been called the “Paradox of
Perception” which defines the way that the problem is initially formulated, and therefore
also influences the development of a solution from a possible range of options (Turton &
Kgathi, 2002). I mporting food that has been grown elsewhere can ameliorate localized
water scarcities. Since it takes 1,000 tons of water to produce one ton of wheat, importing
a ton of wheat effectively imports 1,000 tons of water in an abstract sense. This redefinition can be done without having to admit that water scarcity is a strategically important
factor (hence the importance of being politically silent). But in order to do this, one needs
to redefine political priorities away from national self-sufficiency in food to food security
instead. This move, however, is fraught with political risk, as it opens up a new set of
political and economic dependencies on the developed countries of the North (see Allan,
2000).
• The willingness and ability of all role-players to negotiate in order to generate coping
strategies or develop institutions is an indicator of social ingenuity. I t is the ability to gain
consensus on hydrological data that builds confidence in otherwise-contested river basins.
This also allows the core problem of water-resource management to be re-defined in such
a way that the trade in virtual water can become an effective strategy. The Paradox of
Perception is relevant in this regard (Turton & Kgathi, 2002).
• Finally, the ability of a given social entity to sustain institutions once created is an
indicator of both social and technical ingenuity. This article suggests that countries with a
higher GNP per capita are more likely to sustain institutions than those with a lower GNP
level for a variety of reasons—including the technical ability to generate data on which
incremental decision-making can be based.
66
Po p u l a t i o n a n d W a t e r
relatively high per capita water availability, showing the flaws in merely first-order analyses. Zambia is politically stable but has a low level of economic activity. I t is also negatively affected by the civil wars in both Angola and the DRC. Should Angola, the DRC,
Mozambique, and Zambia manage to solve these problems, they could conceivably become the regional breadbaskets, using their natural resource endowment to balance the
regional water scarcity by becoming virtual water exporters within the Southern African
Development Community (SADC) (Turton et al., 2000b).
The four southern African WP cases present a complex set of problems indeed. Since
there is a relative scarcity of both first- and second-order resources in these cases, their
dependence on external aid is likely to grow over time. Lesotho is an interesting case as
it is first-order resource poor, yet it is also the source of water for South Africa via the
Lesotho Highlands Water Project (LHWP). Water represents one of the few natural resources that Lesotho can exploit (the other being labor and, to a lesser extent,
diamonds). So it sells water to South Africa, using the royalties to finance other development projects. Significantly, all of the East African countries fall into the WP category.
This suggests that East Africa faces relatively more complex development problems than
Southern Africa does.
Som e Hypot heses for Test ing
The results presented in Table 3 suggest a series of hypotheses that can be tested
more exhaustively elsewhere. To review, four such hypotheses are evident:
• I n all cases presented, the relative abundance (or scarcity) of the second-order
resource determines the outcome.
• For countries with a relative abundance of first-order resources and with a relative
scarcity of second-order resources, developmental potential is likely to remain low.
This condition can be labeled Structurally-I nduced Relative Water Scarcity (SI RWS),
an unhealthy condition that policy development should seek to counter vigorously.
• For countries with a relative scarcity of first-order resources and with a relative
abundance of second-order resources, developmental potential is likely to be high.
This condition can be labeled Structurally-I nduced Relative Water Abundance (SI RWA),
a healthy condition to be actively sought as a policy outcome.
• For countries with a relative scarcity of both first- and second-order resources,
developmental potential is likely to remain low. This condition can be labeled Water
Poverty (WP), a debilitating condition that is likely to result in a spiral of social and
economic decay over time, with no apparent end in sight short of external intervention in some form. Under these conditions, policy intervention is likely to be exogenous in nature—dependent on third-party involvement.
I t would be most illuminating to test these hypotheses by means of a more robust
methodology and by using a wider range of indicators. Turton (2002) is developing such
a methodology, along with indicators that are applicable to the management of international river basins. These indicators include aspects such as the ability to generate data
independently of foreign assistance, and the ability to legitimize that data by means of
building consensus among all riparian states. (See Box 2 for more details.) The outcomes
of such a venture would be valuable for policymakers and water-resource professionals in
the developing world.
GI S as a Managem ent Tool—Just a Mat t er of Represent at ion?
The previous detailing of population and water scarcity nuances in the developing
world has laid the groundwork for an assessment of the role of technology in general—
Po p u l a t i o n / W a t e r N e x u s
67
and Geographical I nformation Systems (GI S) in particular—in managing such problems. I s
GI S a helpful tool for gauging population growth and water stress, or is it a manipulative
device for representing the world in the image of the powerful? This is an issue of increasing relevance, meriting far greater attention outside the world of geography and water
resource management. I t is particularly relevant to the developing world.
Since its inception in the developed world in the 1960s, remote sensing has been a
growth industry, becoming a highly popular representational tool to locate three-dimensional data. Yet critics such as Pickles (1991) charge that GI S tends to be used
unreflectively—those who use it are not alert enough either (a) to the assumptions underlying their technology of choice, or (b) the implications of its use. This criticism is
necessarily bound up with value issues and ethics. Like any map, a GI S representation of
the world imposes a set of values on its users. The answer to a research question is
dependent on the assumptions underlying that question. Thus, if the question is whether
GI S can shed light on water and population stress, this not only implies the assumption
that there is a question of stress but also that this stress could lead to problems.
For example, knowledge constructs like “water wars” (most famously coined by Joyce
Starr) and the “population time-bomb” express the pessimistic Malthusian perspective.
These constructs have not gone unchallenged, and as a result the doomsayers seem to
be beating a retreat—see, for example, I CRC (1998), in which Tony Allan argues that it is
the “optimists” who are right (although he deems them dangerous, as they promote
complacency about real challenges to be met). This debate highlights the need to take
solution-capacity into account as much as problem-potential.
I f stress is the ratio of challenge to coping capacity (Lazarus, 1966), then coping with
stress may involve reducing the challenge (needs) or increasing the coping capability
(adaptive capacity). Fortunately, revised projections on population growth and a greater
understanding of virtual water—one example of the adaptive capacity introduced above—
provide a more optimistic view. One such view is Allan’s dictum that the pessimists are
wrong but useful, while the optimists are right but dangerous (Allan, 2000). Researchers
should therefore be careful both to point out what they believe and what information
they rely on to back up those beliefs.
I t is important to realize that GI S is an information management tool rather than a
data-gathering tool. What emerges from a GI S exercise in itself does not say anything
about the policy issue that gave rise to the exercise in the first place. As a consequence,
the “garbage-in, garbage-out” principle applies with a vengeance to GI S. For example, a
researcher might attempt to gauge the world’s level of urbanization by the amount of
light its cities emit. The larger the dots on the world map, the bigger the urban population. Yet this analysis would make sense only if the level of energy use is equal across the
globe, which it obviously is not. There are striking differences between per capita energy
use in Sana’a, Yemen and Cape Town, South Africa; as a consequence, Yemen fails to
appear on some urbanization maps (Allan, personal communication, 2000).
The phrasing of the research question, the data input, and the criteria for assessment all matter, because each impacts on the overall construction of the knowledge that
we seek to build. A good example is early warning in famine policy. I n many emergency
situations, food may well be available, but the mechanisms of exchange (entitlements)
by which people have traditionally gained access to food have ceased to function (Sen,
1981). I n these cases, famine is caused not by a failure of supply but by a failure of
meeting effective demand (Hutchinson, 1998).
The concept of the “water barrier” 3 is a relevant application of these observations on
the nature of questions to be asked to the water sector. The renowned Swedish hydrologist Malin Falkenmark (1990) introduced the concept as a practical rule of thumb, but
eventually she almost came to regret coining it (Falkenmark, personal communication,
68
Po p u l a t i o n a n d W a t e r
1995) as the water sector began to interpret it as an unassailable rule. While the water
barrier is a handy device to show how many countries may be mining their way into
future misery, its subsequent uses ignored other factors such as second-order resources.
The concept of the “water barrier” in itself provides a useful, confrontational view that
underscores an alarmist agenda about the state of the water resource, intended to awaken
governments to the unpleasant realities of current trends in the available water stock.
Falkenmark intended the “water barrier” to provide a guide to the minimum water
requirement for an average human being, which she calculated at 1,700 m3/ cap/ yr -1
(Falkenmark, 1989). The concept was soon enshrined in policy documents as a hard and
fast rule, unfortunately reinforcing existing platitudes that assume water is recovered,
handled, and distributed everywhere in a uniform way, thereby ignoring institutional,
cultural, and economic differences. This problem is common for analyses in which firstorder resources are the sole focus of attention. A meat-eating, industrial-consumer society such as the United States has a rather different water-demand pattern than a vegetarian, self-sufficient nomadic tribe living on a bottle of water a day. Local water scarcity
is also only a problem for an area when non-native people either want to, or have to, live
there. One consequently needs to take into account first-order natural resources, second-order social resources, and the settlement pattern of people if the problem of water
resource availability is to be adequately understood. Taking data as absolutes can easily
lead to non-adaptive conclusions (Geldof, 1994), which are clearly unsatisfactory.
GI S and Social Cont ext
The issue of social context is also critical in appraising the validity of a particular
technological application. Social context suggests (as did the first section of this article)
that there are different types of scarcity. Sexton’s (1992) concept of “economically-induced scarcity” and Warner’s (1992) “politically-induced” scarcity both hint at an underlying mismatch between the water wealth offered by nature and the actual amount of
water available to specific groups and individuals in society. Ohlsson’s (1998, 1999) differentiation between first-order and second-order resource scarcity is also a dynamic
concept; it addresses response to stressors (such as drought, floods, and famine) rather
than viewing scarcity as frozen in absolute terms in a particular moment in time. Countries that are poorly endowed with water resources are not necessarily in trouble if they
have adaptive capacities and mechanisms that are either in place or capable of mobilization before the debilitating effects of absolute scarcity become a limiting factor. A country
that has found ways to use ingenuity—what we would call its “water I Q”— will not always
result in economic stagnation and political instability.
Conversely, a country that is seemingly on the safe side of the “water barrier” does
not necessarily have reasons for complacency. On the basis of this insight, Ohlsson (1998;
1999) has endeavored to rank countries according to proxy indicators of social scarcity,
guided by the UNDP’s Human Development I ndex. As a result, we now have proxy indicators for second-order scarcity that can be developed further if found to be useful—see,
for example, Sullivan et al. (2000). Yet one must bear in mind the “proxy-ness” of the
indicators that are being used in this work. Even if one managed to refine the method to
a high level of mathematical sophistication, there is still the question of reliability of
inputs from official statistics.
I n the above examples of the use of GI S in policy-relevant science, it was the interpretations of geographic information that were at issue rather than the input. Unfortunately, interpretations are in many instances where the trouble begins. I n a country
where an aggressive hydraulic mission is bent on mobilizing more water as a foundation
for socioeconomic development, experts tend to rely on a positivist approach to knowledge —“objective” and precise science, dominated by experts and high technology and
Po p u l a t i o n / W a t e r N e x u s
69
excluding lay knowledge and “fuzziness.” An example is the science of hydrology. While
hydrology claims to be based on hard data and uses mathematical logic, poor data quality in the form of short time-sequences (coupled with the problems of extrapolation)
ultimately yield gross distortions of reality. Another example is flood forecasting, in which
cost-benefit assessments on flood protection are made despite lacking adequate time
series to justify their extrapolations (Green & Warner, 1999). Politically rational processes
are less orderly and predictable than hard science—emotions, values, hard-nosed opportunism get in the way. The measurement of such processes does not lend itself to the use
of concrete numbers, so such processes tend to play havoc with this purported objectivity, with potentially deleterious effects on knowledge-building.
I t should be noted that the GI S experts with whom we have interacted are often well
aware that GI S is not a miracle toolbox but an instrument that necessarily reflects the
biases of its filters and the project’s aims. While the GI S community should perhaps
display even greater receptivity to acknowledging the shortcomings of GI S, it is fair to say
that policymakers are equally responsible when they receive information that reflects
their biases.
The technical ability with which societies are able to handle their water-resource base is
paramount.
Securit izat ion of Wat er
Water resources are often securitized in semi-arid countries, especially when those
countries have a strong hydraulic mission and face closure in shared international river
basins (Turton, 2001). Securitization is the elevation of an issue above normal politics in
light of the perceived national security interest, legitimizing extraordinary management
measures (Buzan et al ., 1998). As a result, in political settings such as I srael, Portugal,
and Turkey, domestic scholars have found it frustratingly difficult to get access to essential water data because the government considers such data too sensitive to release. Like
water, information can become scarce if perceptions of threat to the national interest
prevail. During the Oslo negotiations, Palestinians had to rely on I sraeli water maps that
70
Po p u l a t i o n a n d W a t e r
were open to varied interpretation, thereby undermining confidence in the process. Drawing on Buzan et al’s (1998) work on security strategy, this phenomenon may be called the
securitization of information (Warner, 1998). The securitization of information often makes
official statistics rather dubious, particularly in the Middle East.
When water has been elevated to a national security concern, projects promoting
water development become undebatable. The persistence of this phenomenon has given
rise to a concept known as the “sanctioned discourse,” whereby a select elite determines
what may be said about water-related development projects and by whom. Both in Turkey (the multi-dam Southeast Anatolia Project or GAP, on the rivers Euphrates and Tigris)
and in Egypt (the Tushka project, which seeks to irrigate Egypian desert land by means of
a spillway), criticism of pet projects has been taken as criticism of the state (Warner,
2000).
Abd al-Aziz Ahmad, a senior Egyptian official in the State Hydropower Commission,
generated a series of now-famous reports that raised questions about the long-term
sustainability of the Aswan High Dam (Waterbury, 1979, page 120); he subsequently met
with ostracism in Egypt. Bureaucratic politics (Allison, 1999 [ 1971] ) provide yet another
and especially distorting element in this respect. Large organizations such as governments tend to form what Eric Wolf has labelled the tributary social organization (Wolf,
1990). Government bureaucracies, for example, form a tributary system that collects
resources to finance public works: just as water flows from a tributary stream into a lake,
wealth (through taxes and interest payments) flows towards those bureaucracies or departments that create new projects. Control of water then becomes the institutional
ability to develop new projects (Johnston & Donahue, 1997). Concomitantly, the way to
seek rapid promotion within the dominant organization becomes the completion of a
successful project; and often the grander such projects are, the better they are perceived
by gate-keeping elites. This dynamic favors the alignment and suppression of negative
information; under such circumstances, any statistics generated by powerful water boards
are not necessarily reliable, and consultants are under pressure to conform to the wishes
of the bureaucracy (Brichieri-Colombi, personal communication, 2000).
Thus, as the output of GI S depends both on the input and the questions underlying it,
GI S represents the world in a way that reflects those interests. Depending on what gatekeeping elites want to show, they can manipulate their computer images to highlight and
represent their image of reality. The selection of such filters is a key passage point that is
easily overlooked in evaluating GI S. Commercial applications of GI S can all be used to
accumulate a plethora of personal information building up to detailed demographic profiles, and the intelligence community has always used data from civilian satellite systems
in carrying out its security mission (Morain, 1998). Technologically advanced countries
can spy on less well-endowed countries to see where strategic resources (such as water,
oil, or natural gas) are without them necessarily being able to return the favor.
When we admire GI S data, therefore, we need to consider whose reality the shiny
GI S material represents.
GI S as a Reduct ionist Tool
Apart from the influence of interests, there is a fundamental issue at stake regarding
the way the world is analysed. At the philosophical, epistemological, and ontological
level, GI S represents a Western tradition of decomposing the world into minutiae rather
than integrating it holistically. Conventions such as the Convention on Climate Change
require countries to collect data according to a certain model that (a) brackets in countables
prioritized by the developed world, and (b) brackets out non-countables such as cultural
and religious values as well as common-law institutions. The unified approach on measuring vulnerability to climate change proved much more compatible with countries like
Po p u l a t i o n / W a t e r N e x u s
71
the Netherlands than islands in the South Pacific, which not only find it harder to come by
physical and economic data to fit into the system, but also have raised the issue of
different norms such as communal ownership of resources. Forcing them to adopt a
framework reflecting Western-oriented values may well mean the non-appearance of
local values in the comparative data, which will make it harder for them to bargain for
specific compensatory and rehabilitating measures.
Also, just as school atlases of yore listed each colony’s raw materials, GI S can be
used to pinpoint the last exploitable energy and water reserves and, in so doing, promote
their capture. A transferred technology may not conform to indigenous values held by
beneficiaries. This disconnect becomes ever more salient as GI S technologies develop
into a must-have for the developing South. To keep up, the developing countries seek
access to the same reality-transforming technology that they see in the developed world.
The use of GI S is now universal in the West, but close allies such as Egypt also benefit
from the latest American technology. Yet poorer countries, or those less sympathetic to
Western-style capitalism, may be left out: for example, peoples without officially-recognized states (such as the Palestinians) cannot legally buy remote-sensing images, which
puts them at a distinct disadvantage. The same issue of availability and access is consequently not just relevant for water, but also for information tools like GI S.
There is also the issue of whether a country that is the subject of GI S analysis can
actually do anything with the results that the technology yields in the first place. When
floods hit Mozambique in February 2000, flood-warning technology was well in place in
the region, but Mozambique did not seem to benefit from it for lack of a working knowledge infrastructure. Thus, no matter how promising epistemic communities are (such as
those that involved in global warming and regional-knowledge building), strictures such
as those of communication, representation, and logistics may prevent the development
of an effective regional water security regime for Southern Africa.
GI S as a Managem ent Tool
Thus far, this article has mainly pointed out serious questions about much of the
current use of GI S. But we should not overlook the possibilities GI S opens up if it is used
mindfully. The aim should be to make GI S a more varied toolbox for understanding the
world. The technology lends itself to control and resistance strategies—both between
countries and between state and society. Like the I nternet, the increasing affordability of
GI S could, with time, democratize the technology into a tool for use by those with only a
reasonably powerful computer. From controversial dams in I ndia to river-deepening and
widening projects in the Netherlands, from illegal industrial pollution to covert military
operations, researchers who pool resources should be able to parse potentially distorted
official statistics and get the “real” numbers. And, in this regard, GI S could promote the
democratization of science for policy in general. Fortunately, the ever-falling cost of GI S
also allows previously-silenced critics to turn the tables and break the monopoly of information use by elites.
What Turton & Ohlsson (1999) have called the “second transformation” starts with
new voices on the scene, such as civil-society groups consisting of NGOs and communitybased organizations. While underestimated at first as being irrelevant and repressed,
these voices have increasingly started to be integrated into government, particularly
where institutional government capacity is weak. I n this role, such voices provide a chance
to mobilize the second-order resources that some countries may lack by mobilizing untapped knowledge and governance capacities and bridging the ingenuity gaps. The social
and environmental consciousness that these advocates express has started to challenge
the approach of those still enthralled by a “hydraulic mission” (Reissner, 1996).
I n the light of these developments, hailing GI S as the return of positivism would be to
72
Po p u l a t i o n a n d W a t e r
paint too simplistic a picture of a changing reality. I t would make more sense to try and
convince those who work with GI S that theirs is one knowledge among many—such as
the traditional knowledge systems of many local water users in the developing world.
I nterestingly, noted natural scientists for some time now have been advocating the involvement of non-experts in policy debates to help decide on contested value-laden policy problems as well
as those surrounded by a great measure of uncertainty (e.g., Funtowicz & Ravetz, 1983). BringWh e n s e c o n d -o r der
ing in “lay” (non-expert) voices and rationalities, perceptions, and emotions as considerresourc es are
ations for policymaking (Geldof, 1994) will be
m o b i l i ze d i n
perceived by some as a striking blow to the
s
u
f
f
i
c ient quant it ies
positivist outlook. However, this article advoand in suffic ient t im e,
cates: (a) promoting the adaptation of GI S
t he pit falls of rapid
systems such that they allow for a diversity of
questions to be raised; and (b) making GI S a
populat ion grow t h
tool that can also be operated by those of limc a n b e ave r t e d .
ited means or those seeking to promote an alternative, counter-hegemonic agenda. The difference between availability and actual access is also
a crucial one. GI S can be especially helpful in showing
not just the location and distribution of people, but by showing the physical infrastructure or “pipelines of power” (Turton, 2000a), thus showing how
hydraulic structures can be developed to ensure differential access to water.
But the mainstream GI S community is confused by these criticisms, and dialogue
towards progress on these issues has so far been painful and generally non-productive.
As Schuurman (2000) notes, GI S experts have problems coming to terms with the language of GI S critics. Social science can also do its bit by phrasing its arguments in language that is intelligible to those who have been trained in the natural sciences. Fortunately, a new generation of engineers and physical geographers seem to be more sensitized to these questions than their predecessors were. One example is I nitiative No. 19 of
the University of California-Santa Barbara’s National Center for Geographic I nformation
Analysis’ (NCGI A) I nitiative, in which critics of GI S work together with their GI S-savvy
peers (Schuurman, 2000). But we will need to do much more—ultimately redesigning
engineering, geography, and social science curricula in a cross-disciplinary way so that
the next generation will learn to speak multidisciplinary languages understandable to a
wider audience.
Key Quest ions
Despite the debate over the values and shortcomings of GI S, it remains an important
tool in the water availability/ scarcity debate. With that debate in mind, it is now possible
to focus attention on answering four critical questions.
Quest ion 1. Will t here be enough w at er t o support regional populat ions in t he
fut ure?
The African cases presented (even those characterized here as “low”) almost all
show an alarmingly high rate of population growth when compared to trends in the
developing world. The doubling and even tripling of populations over the 39-year period
for which data have been selected is cause for alarm. I n terms of first-order analysis
alone, this phenomena represents a significant reduction in the availability of water per
Po p u l a t i o n / W a t e r N e x u s
73
capita over time—ranging from half to a third over the period. When second-order resources are mobilized in sufficient quantities and in sufficient time, however, the pitfalls
of rapid population growth can be averted. Second-order resource management therefore becomes the key management issue, relevant to water resource managers, aid
agencies, and foreign policy practitioners alike.
SI RWA countries have a wider range of options available to them and are likely to
manage water scarcity more effectively than SI RWS countries. SI RWA countries have the
problem of mobilizing more water, so the issue of “running out of water” (another flawed
concept that is often used in first-order analysis) is more relevant to them; but given their
capacity to adapt, they are likely to implement coping strategies in time to avert a disaster. Virtual water trade is likely to become more important for these countries, raising the
issue of increased vulnerability to global grain price fluctuations, increased dependence
on erstwhile colonial powers, and other strategic considerations. SI RWS countries, in
contrast, do have the problem of developing the water resources that they naturally
have. WP countries are likely to face catastrophe after catastrophe with crisis management being the norm, so they are less likely to maintain social, economic, and political
stability. Water scarcity is therefore likely to become a critical developmental constraint,
with its debilitating effects unevenly distributed within WP countries and potentially exported regionally in a domino-effect of instability.
Quest ion 2. Can Geographical I nform at ion Syst em s ( GI S) t echnology be used
t o m ap w at er resources and fut ure populat ion grow t h?
Clearly GI S is a powerful management tool with enormous potential. There are a
number of pitfalls, however, as discussed above. First, as noted in the introduction, political legitimacy and accountability are generally low in the developing world. Under such
conditions, resource capture by the economic elite is increasingly likely. A powerful tool
like GI S can therefore become an instrument of manipulation and political control rather
than a water-management support platform. The impact of this should not be underestimated.
Second, while GI S represents an information management tool, it is not a science
(Wright et al., 1998). As such, its effectiveness is hampered by the type and quality of
data that originally available for input. I n SI RWA countries, the likelihood of adequate
primary data (coupled with the existence of sufficient intellectual capital and institutional
capacity with which to collect, store, process, interpret, and share that data) is such as to
generate optimism about GI S’s applicability. For those countries, GI S is thus likely to
become a powerful management tool in the future, and in many cases this trend is
already evident. For SI RWS countries, the lack of substantial second-order resources is
likely to mean that institutional development will be low and intellectual capital will be
scarce; as such, the prognosis for the success of GI S in these cases is dubious. The same
holds true for WP countries.
Third, the issue of North/ South dependency becomes relevant. I n the case of GI S,
the technology is developed in the industrialized North and selectively exported to the
developing South, possibly exacerbating the existing maldistribution of global power and
creating new forms of marginalization and dependency.
Quest ion 3. Has t he quest ion now becom e one of m anaging dem and for w at er
rat her t han supply?
There is no simple answer to this question because it is dependent on a series of
other issues—such as (a) the capacity of a state to negotiate with riparian states in
shared river basins, along with (b) the ability to develop the institutional capacity neces-
74
Po p u l a t i o n a n d W a t e r
sary for developing effective coping strategies. I n this regard, the Turton-Ohlsson typology presented in this article provides a useful framework for unpacking these issues.
There are at least three necessary conditions for demand management to succeed:
• There must be sufficient institutional, intellectual, and administrative capacity in
order to generate viable water-demand-management (WDM) solutions in the first
place. Similarly, there must be the capacity to meter water consumption, bill users
accordingly, collect payment, and sanction those who do not pay. All of these aspects
are second-order-resource in orientation.
• There must be a high level of political legitimacy if WDM policies are to be supported by the general public.
• There must be a culture of payment for water services received. This also implies
that there must be general acceptance of water as an economic resource.
For SI RWA countries, the prime management issue is about doing more with less.
Under these conditions, WDM is likely to become an important component of a management strategy; however, as Gilham & Haynes (1999) have demonstrated, it is unlikely to
be the sole solution. Where WDM is implemented, political legitimacy is likely to be
severely tested. I n Zambia and in Botswana, for example, attempts by the government to
introduce charges for water (which was culturally seen to be a gift from God) are placing
strains on the political system. The challenge under these conditions is therefore to develop the right mix of culturally-appropriate and politically-acceptable supply- and demand-sided solutions. Supply management will always remain important, with a shift of
emphasis away from large water-transfer schemes to more sophisticated desalinization
and water recycling systems. I t is therefore simplistic to assume that a hard transition will
occur from supply-sided to demand-sided management. I n reality, both elements are
needed in an effective water-resource policy, but with a shift in emphasis between the
two over time toward the management of demand.
The second-order scarcities in SI RWS countries are likely to inhibit the development
and implementation of viable WDM strategies. Popular support for WDM is unlikely in
these cases, and civil disobedience can be expected to actively undermine such policies.
The lack of institutional capacity in SI RWS cases is also likely to mean that water meters
are not installed, billing capacity is likely to be non-existent, and legal sanction for noncompliance lacking. Clearly under such conditions the prognosis for success is low. The
same holds true for WP countries.
Quest ion 4. How w ill WDM be achieved?
Since the key management issue revolves around second-order resources, three elements are likely to be crucial:
• On the structural side, institutional development is important. Such institutions
should: (a) be adequately staffed; (b) have sufficient data processing and sharing
capabilities with which to develop and monitor solutions; and (c) be adequately funded
in order to ensure sustainability. I t is critical for demand-management institutions to
meter water consumption, generate bills and collect monies due, and prosecute those
who adopt a non-compliant posture.
• On the social side, there needs to be a culture of payment and a high level of
support for the decision-making bodies. A cultural acceptance of water as an economic resource that has value and should be paid for is critical.
• On the political side, there needs to be unconditional support by politicians. Without this support, the overall credibility of the management process will be underPo p u l a t i o n / W a t e r N e x u s
75
mined. I f politicians continue to promise free water to potential voters, then WDM
strategies will be compromised.
Current research underway at the African Water I ssues Research Unit suggests that
three components are necessary to manage water demand, at least in an African context. The first of these is accessibility to water . Where water is inaccessible, its use is low
and the time taken to fetch it is high. These dynamics change when water becomes more
readily available and convenient to use. This means that the second component of any
given demand management strategy is pricing. As water becomes more readily available,
people are willing to pay for the resource. Demand can be managed through an innovative tariff structure such as that currently used in Durban, South Africa—but this is only
effective if adequate access to water has already been established and if people’s attitudes to the use of water have changed. The third component is, consequently, education. Education must target a wide spectrum of audiences—from water users up through
the water supply chain to the political level. I f politicians continue to offer free water as a
means of securing votes, demand management is doomed to fail! An important end-goal
of the education process is to change the attitude that water is a free good, in keeping
with the Dublin Principles (I CWE, 1992) and World Water Vision (Cosgrove & Rijsberman,
2000).
Conclusion
The development and sustainability of second-order resources determine how well a
society can manage a resource such as water. Typically, this type of resource is in short
supply in the developing world. Hydropolitically-related foreign policy initiatives are likely
to fail if this subtle but important nuance is not taken into consideration. Many cases of
aid dependence result directly from an attempt to stimulate development in the absence
of any recognition of the importance of second-order resources. Similarly, applications of
modern technology such as GI S is likely to fail if second-order resources are not taken
into account. Where correctly applied, however, GI S is likely to become a powerful and
equalizing management tool of the future. The strategic significance of some of these
nuances is important, given the impact of the 2001 terror attacks on New York’s World
Trade Center and the I ndian Parliament. The foundation of this strategic significance
derives from the fact that there is a correlation between (a) countries that have the
potential to export terror, and (b) the existence of WP as defined in this article.
Ac k now ledgem ent s
The authors wish to acknowledge the valuable support by the Woodrow Wilson I nternational Center for Scholars for sponsoring this project and enabling it to be presented to
a wider audience. I n particular, this support enabled Richard Meissner, a research associate at the African Water I ssues Research Unit (AWI RU) based at the Centre for I nternational Political Studies (CI PS) at Pretoria University, to spend some time in London collecting data and consulting with Jeroen Warner. The Water Research Fund for Southern
Africa (WARFSA) is also acknowledged for its support to the current WDM study that is
being conducted at AWI RU. However, the authors alone accept the responsibility for the
conclusions of the project.
Not es
1
This distinction is not a clinical one, however, because many other criteria could be used. Even in
this case, there are still overlaps. Tanzania, for example, falls into both classifications.
76
Po p u l a t i o n a n d W a t e r
2
The “water wars” argument suggests that, as a country’s uncontrolled population growth erodes
its available water resources, conflict potential in that country will increase to the point where war
over water is inevitable (Turton, 2000b, page 39). By relying on so-called “hard” primary data
(population and water availability), this linkage results ultimately in a teleological argument. I n
reality, this so-called “hard” data are not hard at all; it involves a high level of generalization
combined with specific assumptions. For example, U.S. Census (2000) lists Angola’s population in
2000 at 10,145,000, while the UN (2000) World Population Data reports a figure of 13,134,000. At
best, such data are broad generalizations only and should not be regarded as being the final word
on the issue.
3
The “water barrier” was defined by Falkenmark (1990:181) as a conceptual “barrier” that was set
at 2,000 people per standard “flow unit,” consisting of one million cubic meters of water per year.
Falkenmark considered any figure above the water barrier to make any form of economic development virtually impossible given current technologies.
Subsequent analyses have shown difficulties in universal application of the water barrier.
I srael, for example, seems to be capable of surviving at a figure well beyond that set by
Falkenmark. South Africa is approaching the barrier and also seems set to survive the transition.
These anomalies have given rise to new explanations, leading to the concept of second-order
resources. I n the cases where states can survive beyond the water barrier, they all have high
levels of second-order resources.
Re fe r e n c e s
Abrams, Len J. (1996). “Policy development in the water sector—The South African experience.” I n
P. Howsam & R.C. Carter (Eds.), Water policy: Allocation and management in practice (pages
21-30) . London: E & FN Spon.
Allan, J.A. (1996). “Water use and development in arid regions: Environment, economic development and water resource politics and policy.” Water Use and Development 5 (2), 107-115.
Allan, J.A. (2000). The Middle East water question: Hydropolitics and the global economy. London:
I B Tauris.
Allison, Graham. (1999, 1971). Essence of decision . Boston: Little, Brown.
Beck, Ulrick. (1992). Risk society—Towards a new modernity . London: Sage.
Bradnock, Robert W. & Saunders, Pat L. (2000). “Sea-level rise and submergence. The political
ecology of environmental change in the Bengal delta.” I n P. Stott & S. Sullivan (Eds.), Political
ecology: Science, myth and power (pages 66-90). London: Edward Arnold.
Buzan, Barry; Waever, Ole & De Wilde, Jaap. (1998). Security: A new framework for analysis.
London: Lynne Rienner.
Christie, Francis & Hanlon, Joseph. (2001) Mozambique and the great flood of 2000. Oxford:
James Currey. I ndianapolis: I ndiana University Press.
Cosgrove, William & Rijsberman, Frank. (Eds.) (2000). World water vision: Making water everybody’s
business. London: Earthscan.
Fairclough, Norman. (1989). Language and power. London & New York: Longman.
Falkenmark, Malin. (1986). “Fresh water: Time for a modified approach.” Ambio 15 (4);
194-200.
Po p u l a t i o n / W a t e r N e x u s
77
Falkenmark, Malin. (1989). “The massive water scarcity now threatening Africa: Why isn’t it being
addressed?” Ambio 18 (2), 112-118.
Falkenmark, Malin. (1990). “Global water issues confronting humanity.” Journal of Peace Research
27 (2), 177-190.
Falkenmark, Malin & Lundqvist, Jan. (1995). “Looming water crisis: New approaches are inevitable.” I n L. Ohlsson (Ed.), Hydropolitics: Conflicts over water as a development constraint
(pages 178-212). London/ New Jersey: Zed Books.
FAO. (2000). The Food and Agricultural Organization of the United Nations. Population: Annual
time series. [ On-line] . Available: http:/ / apps.fao.org/ page/ collections?subset= agriculture
Funtowicz, Silvio & Ravetz, Jerry. (1983). “Scientific predictions and the methods of technological/
environmental assessment.” SERC-SSRC Risks Project, Dept of Philosophy, Leeds University,
Oct 1983.
Geldof, G.D. (1994). Adaptief waterbeheer (in Dutch). Deventer: Tauw Civiel en Bouw.
Gilham, Steven & Haynes, Michael. (1999). Demand management: A possible alternative to augmentation? A South African case study (AWI RU Occasional Paper). [ On-line] . Available at
http:/ / www.up.ac.za/ academic/ libarts/ polsci/ awiru
Gleick, Peter (ed.). (1993). Water in crisis: A guide to the world’s freshwater resources. New York:
Oxford University Press.
Haas, Peter M. (1993). “Epistemic communities and the dynamics of international environmental
co-operation.” I n Volker Rittberger (Ed.), Regime theory and international relations (pages
168-202). Oxford: Clarendon Press.
Homer-Dixon, Thomas F. (1994). “Environmental scarcities and violent conflict: Evidence from
cases.” I nternational Security 19 (1), 5-40.
Homer-Dixon, Thomas F. (1995). “The ingenuity gap: Can poor countries adapt to resource scarcity?” Population and Development 21 (3), 587-612.
Homer-Dixon, Thomas F. (1996). “Environmental scarcity, mass violence and the limits to ingenuity.” Current History 95, 359-365.
Homer-Dixon, Thomas F. & Percival, Valery. (1996). Environmental scarcity and violent conflict:
Briefing book. Washington, DC: American Association for the Advancement of Science
Homer-Dixon, Thomas. (2000). The ingenuity gap. London: Jonathan Cape.
Hutchinson, Charles F. (1998). “Social science and remote sensing in famine early warning.” I n
Diana Liverman (Ed.), People and pixels: Linking remote sensing and social science (pages
189-196). Washington, DC: National Academy Press.
I nternational Conference on Water and Environment (I CWE). (1992). The Dublin Statement on
water and sustainable development . Dublin, I reland: I nternational Conference on Water and
Environment. [ On-line] . Available: http:/ / www.wmo.ch/ web/ homs/ icwedece.html
Johnston, Barbara Rose & Donahue, John M. (1998). Water, culture & power: Local struggles in a
global context . Washington, DC: I sland Press.
Kaplan, Robert D. (1994, February). “The coming anarchy.” Atlantic Monthly , 44-76.
78
Po p u l a t i o n a n d W a t e r
Lazarus, Richard S. (1966). Psychological stress and the coping process. New York: Academic
Press.
Liverman, Diana. (1998). People and pixels: Linking remote sensing and social science. National
Research Council. Washington, DC: National Academy Press.
Luke, Timothy W. (1991). “The discipline of security studies and the codes of containment.” Alternatives 16 (2), 315-344.
Lundqvist, Jan. (1998). “Avert looming hydrocide.” Ambio. 27 (6), 428-433.
Moran, Emilio F. (1998). “A brief history of remote sensing applications, with emphasis on Landsat.”
I n D. Liverman (Ed.), People and pixels: Linking remote sensing and social science . Washington, DC: National Academy Press.
Ohlsson, Leif. (1998). Water and social resource scarcity—An issue paper commissioned by FAO/
AGLW. Presented as a discussion paper for the 2nd FAO E-mail Conference on Managing Water
Scarcity. WATSCAR 2.
Ohlsson, Leif. (1999). Environment, scarcity and conflict: A study of Malthusian concerns. Department of Peace and Development Research. University of Göteborg, Sweden.
Ohlsson, Leif & Turton, Anthony R. (1999). The turning of a screw . Paper presented in the Plenary
Session of the 9th Stockholm Water Symposium “Urban Stability through I ntegrated WaterRelated Management,” hosted on 9-12 August by the Stockholm Water I nstitute (SI WI ) in
Stockholm, Sweden. [ On-line] . Available: http: / / www.soas.ac.uk/ Geography/ WaterI ssues/
OccasionalPapers/ home.html
Ohlsson, Leif (with Lundqvist, Jan). (2000). “The turning of a screw—Social adaptation to water
scarcity.” I n Malin Falkenmark et al., New dimensions in water security (Part 3) . Rome: FAO/
AGLW.
Ó Tuathail, Gearoid; Dalby, Simon; & Routledge, Paul. (Eds.). (1998). The geopolitics reader . London: Routledge.
Pickles, John (Ed.). (1991). Ground truth . New York: Basic Books.
Prunier, R. (1995). The Rwanda crisis: History of a genocide . New York: Columbia University Press.
Reissner, Mark. (1993). Cadillac desert: The American west and its disappearing water . Revised
Edition. New York: Penguin.
Rindfuss, Ronald R. & Stern, Paul C. (1998). “Linking remote sensing and social science: The need
and the challenges.” I n D. Liverman (Ed.), People and pixels. Linking remote sensing and social
science. National Research Council. Washington, DC: National Academy Press.
Schuurman, Nadine. (2000). “Trouble in the heartland: GI S and its critics in the 1990s.” Progress in
Human Geography 24 (4), 569-590.
Sen, Amartya. (1982). Choice, welfare and measurement . Oxford: Blackwell.
Sexton, Robert. (1992). “The Middle East water crisis: I s it the making of a new Middle East
regional order?” Capitalism, Nature, Socialism 3 (4), 65-77.
Solomon, Hussein & Turton, Anthony (Eds.). (2000). Water wars: An enduring myth or impending
reality? African Dialogue Monograph Series No. 2. Durban: Accord Publishers.
Po p u l a t i o n / W a t e r N e x u s
79
Starr, Joyce. (1991, Spring). “Water wars.” Foreign Policy 82, 17-36.
Stott, Philip & Sullivan, Sian (Eds.). (2000). Political ecology: Science, myth and power . London:
Edward Arnold.
Sullivan, Caroline; Meigh, J.; Simon, S.; Lawrence, P.; Calow, R.; & McKenzie, A. (2000). The
development of a water poverty index: A feasibility study (Final report prepared for DFI D).
Wallingford, UK: Centre for Ecology and Hydrology.
Turton, Anthony. (1999). Water scarcity and social adaptive capacity: Towards an understanding of
the social dynamics of managing water scarcity in developing countries (MEWREW Occasional
Paper No. 9, SOAS Water I ssues Study Group). [ On-line] . Available: http: / / www.soas.ac.uk/
Geography/ WaterI ssues/ OccasionalPapers/ home.html
Turton, Anthony. (2000a). “Precipitation, people, pipelines and power: Towards a political ecology
discourse of water in Southern Africa.” I n P. Stott & S. Sullivan (Eds.), Political ecology: Science, myth and power (pages 132-153). London: Edward Arnold.
Turton, Anthony. (2000b). “Water wars in Southern Africa: Challenging conventional wisdom.” I n
H. Solomon & A.R. Turton (Eds.), Water wars: An enduring myth or impending reality? African
Dialogue Monograph Series No. 2 (pages 35-63). Durban: Accord Publishers.
Turton, Anthony. (2002). “The political aspects of institutional developments in the water sector:
South Africa and its international river basins.” Draft copy of a D.Phil. Dissertation, Department
of Political Sciences, University of Pretoria.
Turton, Anthony. (Forthcoming). “The political dynamics of institutional development in contested
international river basins: A perspective from the south.” [ On-line] . Available in draft form:
http:/ / www.up.ac.za/ academic/ libarts/ polsci/ awiru
Turton, Anthony & Kgathi, Donald. (2002). “The problematique of WDM as a concept and as a
policy: Towards the development of set of guidelines for Southern Africa.” I n I UCN, Water
demand management program for southern africa: Phase I I . Harare: I UCN (ROSA).
Turton, Anthony & Ohlsson, Leif. (1999). Water scarcity and social adaptive capacity: Towards an
understanding of the social dynamics of managing water scarcity in developing countries .
Paper presented in the Workshop No. 4: Water and Social Stability of the 9 th Stockholm Water
Symposium “Urban Stability through I ntegrated Water-Related Management,” 9-12 August.
St ockholm, Sw eden: St ockholm Wat er I nst it ut e ( SI WI ) . [ On-line] . Available: ht t p: / /
www.soas.ac.uk/ Geography/ WaterI ssues/ OccasionalPapers/ home.html
Turton, Anthony; Khupe, John.; & Mucheleng’anga, Chitaku. (2000a). “Research project on water
demand management, natural resource reconstruction and traditional value systems: What
are we learning so far?” I n Proceedings of the WARFSA/ WaterNet Symposium, held in Maputo
1-2 November 2000. [ On-line] . Available: http:/ / www.up.ac.za/ academic/ libarts/ polsci/ awiru
Turton, Anthony; Moodley, Shomenthree; Goldblatt, Michael & Meissner, Richard. (2000b). An analysis
of the role of virtual water in Southern Africa in meeting water scarcity: An applied research
and capacity building project. Johannesburg: Group for Environmental Monitoring (GEM). [ Online] . Available: http:/ / www.up.ac.za/ academic/ libarts/ polsci/ awiru
United Nations (UN). (2000). World population prospects: The 2000 revision . [ On-line] . Available:
http:/ / www.un.org/ popin/
U.S. Census. (2000). U.S. Census Bureau I DP Summary Demographic Data. [ On-line] . Available:
http:/ / www.census.gov/ ftp/ pub/ ipc/ www/ idbsum.html
80
Po p u l a t i o n a n d W a t e r
Warner, Jeroen. (1992). The politics of diversion. Bridging troubled waters in the Middle East .
Unpublished MA thesis, University of Amsterdam.
Warner, Jeroen. (1993). Tightropes across the river. Managing conflict in Middle East river basins
(Occasional Paper, No. 17). Amsterdam: Middle East Research Associates.
Warner, Jeroen. (1998). Post-Westphalian water security. Unpublished paper presented at a conference on occasion of the 350 th anniversary of the Peace of Westphalia, University of Twente,
July 1998. Authors.
Warner, Jeroen. (2000). I mages of water security: A more integrated perspective. AWI RU Occasional Paper No. 3. [ On-line] . Available: http:/ / www.up.ac.za/ academic/ libarts/ polsci/ awiru
Waterbury, John. (1979). Hydropolitics of the Nile Valley. New York: Syracuse University Press.
Weber, Max. (1947). Max Weber: The theory of social and economic organization. New York:
Oxford University Press.
Wolf, Eric (1990). “Distinguished lecture: Facing power—old insights, new questions.” American
Anthropologist 92, 586-596.
World Bank. (1999). Developing countries—Water supply and sanitation—1999. [ On-line] . Available: http:/ / devdata.worldbank.org/ exte…comp.asp?rmdk= 110&smdk= 473898&w= 0
World Bank. (2000). World Bank atlas 2000. Washington, D.C.: The World Bank.
Wright, D. J.; Goodchild, M.F; & Proctor, J.D. (1997). “Demystifying the persistent ambiguity of GI S
as ‘tool’ versus ‘science.’” The Annals of the Association of American Geographers 87 (2), 346362. [ On-line] . Available: http:/ / dusk.geo.orst.edu/ annals.html
Po p u l a t i o n / W a t e r N e x u s
81
ECSP Pu b l i c at i o n s Ava i l a b le
♦ Environmental Change and Security Project Report , I ssues 1-7
(published annually)
♦ China Environment Series, I ssues 1-4 (published annually)
♦ Climate Action in the United States and China (policy brochure)
♦ AVI SO (public policy briefing series, I ssues 1-8 currently available)
♦ PECS News (bi-annual newsletter)
♦ Hong Kong Green NGOs and Environmental Journalism Forum
♦ Protecting Regional Seas: Developing Capacity and Fostering
Environmental Cooperation in Europe (conference proceeding
volume)
♦ The WTO and MEAs: How Should the World Trade Organization
and the MEAs I nterface? (from the Trade and Environment
Forum at http:/ / wwics.si.edu/ tef/ index.htm)
♦ Environmental Change and Security I nterviews (audio compact
disc compilation)
♦ “Contagion and Stability: I mplications” (a world wide web policy
brief: http:/ / ecsp.si.edu/ simulation.htm)
♦ “The Future of the US-Mexico Border: Population, Development,
Water” (a world wide web workshop report: http:/ / ecsp.si.edu/
tijuana.htm)
I f you are interested in obtaining copies of any of these publications, please
contact the Project at: Environmental Change and Security Project, Woodrow
Wilson I nternational Center for Scholars, One Woodrow Wilson Plaza, 1300
Pennsylvania Avenue, NW, Washington, DC, 20004-3027; Tel: 202-6914130; E-mail: ecspwwic@wwic.si.edu.
82
Po p u l a t i o n a n d W a t e r