Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin

Water Resources and Economics ] (]]]]) ]]]–]]]

Contents lists available at ScienceDirect

Water Resources and Economics journal homepage: www.elsevier.com/locate/wre

Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin Djiby Racine Thiam a,n, Edwin Muchapondwa b, Johann Kirsten c, Magalie Bourblanc c a b c

Center for Development Research (ZEF), Walter-Flex Strasse 3, University of Bonn, Germany Environmental Economics Policy Research Unit, School of Economics, University of Cape Town, South Africa Department of Agricultural Economics, Extension and Rural Development, University of Pretoria, South Africa

a r t i c l e in f o

a b s t r a c t

Article history: Received 27 May 2014 Received in revised form 29 October 2014 Accepted 4 November 2014

This paper uses the water-reallocation scheme created within the National Water Act (1998) to analyze the impacts of water policy on farm livelihoods in South Africa. Based on one of the most water stressed catchments in the country, the Olifants river basin, we provide an integrated modeling approach combining water and agricultural modules to investigate the impacts of compulsory licensing and water market on crop production and investment made to improve water use efficiency. The model maximizes net farm profits and takes into account the characteristics of the agricultural sector in the region in classifying farmers between large-scale (LSFs) and emerging (EFs) groups, according to their land acreage, irrigation efficiency and historical heritage. Compulsory licensing is analyzed through curtailment of water-use rights from large-scale to emerging farmers. The water market is investigated to provide conditions under which farms trade water to complete their irrigation schedules. Our results show that, though compulsory licensing might promote a rise in emerging farmers and a re-balance of past riparian-based water allocation schemes, care should be given to the level of that curtailment rate in order to balance equity measures with efficiency objectives. Indeed, we found that the losses associated with water curtailment for LSFs are not entirely captured by the EFs. Therefore, beyond water policy, there are other factors, which also influence farms’

Keywords: Water policy South Africa Olifants river basin Agriculture Water rights Integrated modeling approach

n

Corresponding author. Tel.: þ49 228 73 49 02. E-mail address: [email protected] (D.R. Thiam).

http://dx.doi.org/10.1016/j.wre.2014.11.001 2212-4284/& 2014 Elsevier B.V. All rights reserved.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

2

profits and water use efficiency. The results also demonstrate that a water market provides a good opportunity to increase water use efficiency. The introduction of water market induces LSFs with good water storage facilities a possibility to trade their remaining water-use rights. It also offers EFs an alternative to diversify their water supply sources when they encounter shortfalls in amount of water allocated. & 2014 Elsevier B.V. All rights reserved.

1. Introduction Water allocation in South Africa is undergoing many changes to integrate additional aspects such as equity in distribution among the different users, control of the resource’s sustainability and integration of local stakeholders into the water management practices. High water demand for food production and industry development has stressed many catchments and deteriorated water quality in South Africa [1–8]. Moreover water allocation for agricultural use, the major water user in the country, used to be tied to land ownership hence excluding historically disadvantaged individuals1 from access to water rights. This paper uses one of the most water stressed catchment in the country – the Olifants river basin – to assess the impacts of water market (WM) reform and the policy of compulsory licensing (CL) on farms’ livelihoods. These policies are chosen because they are among the main policy incentives incorporated in the National Water Act (1998) and the National Water Services (1997) to improve water allocation for productive purposes2 within the country. For instance, compulsory licensing is a policy which aims at promoting a re-allocation of water resources in water stressed3 catchments in South Africa. Beyond areas already under water stress, compulsory licensing is also applied to areas where water stress is expected and water quality is damaged by pollution. The water market is a mechanism used in promoting a voluntary transfer of water-use rights for financial compensation. The water market provides the possibility of water trade between farmers, after the public authority has already allocated the water-use-rights. In the agricultural sector water market assumes that farms holding licenses that are not used after a completion of irrigation schedules (surplus license holders) sell such licenses to the ones that still need additional water (deficit license holders) to complement their irrigation schedules. The South African water sector has experienced many institutional and policy reforms, since the democratization of the nation in 1994. Major elements of the water sector reforms included removal of price subsidies, compulsory licensing, and promotion of water trade (water market) to improve efficiency in water use and allocation [10–13]. Such reforms have also established a new institutional structure (e.g. catchment management agencies-CMA and water user associations-WUA) that promotes a more inclusive water management practice. However, despite these reforms little effort is made in investigating how such policy changes might influence farmers’ production and investment decisions in the agricultural sector. This is what the present study aims to contribute to in assessing the impacts of compulsory licensing and water trade on crop production choice and investment made to improve water use efficiency. To the best of our knowledge, no empirical study has computed the impacts of CL on the agricultural sector in South

1 In the Preferential Procurement Policy Framework Act, Historically Disadvantaged Individual (HDI)” refers to South African citizen—(a) who, due to the former apartheid policy, had no franchise in national elections, prior to the introduction of the Constitution of the Republic of South Africa, 1983 (Act No 110 of 1983) or the Constitution of the Republic of South Africa, 1993 (Act No 200 of 1993) and/or (b) who is a female; and/or (c) who has a disability. 2 To direct water to its most economically productive uses and induce adoption of more efficient water use technologies and conservation practices. 3 It is important to highlight that water scarcity (stress) is different from vulnerability in water access. Different indicators (indexes) are provided for a determination of water scarcity (i.e: Falkenmark indicator, Basic Human Water Requirement, Social Water Stress index etc). [9] provides the panorama of indicators measuring both water stress and water access vulnerability.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

3

Africa. With regards to water market (WM), few studies have already provided advantages and drawbacks of such policy on the agricultural sector [14–16], but none have clearly stated under what condition WM might provide benefits to farmers, especially the emerging ones who rely heavily on water allocated by the public authority for irrigation. This study is relevant to the water and agricultural policymakers, since it outlines the likely effects of compulsory licensing and water market reforms on farm livelihoods. Since both policies are not yet implemented in the region this work provides guidelines to policymakers to evaluate the effects that these targeted policies might have generated in case of implementation. In assessing the impacts of CL and WM on farms’ livelihoods, we build an integrated optimization model combining agricultural as well as water modules. The agricultural module consists of representative farms using water for irrigation purposes and producing crops that are sold at a market price. Farms are chosen while considering heterogeneity in land acreage, irrigation efficiency and past historical heritage. In doing so, they are differentiated between large-scale (LSFs) and emerging (EFs)4 groups. LSFs hold more land acreage and use more efficient irrigation technology than EFs, which hold small areas of land and use less efficient irrigation technology. The agricultural module investigates how the selected water policies impact on crop production, water used in irrigation and investment made for an increase in irrigation efficiency. The water module considers water allocated by public authorities to farmers according to the water policies. The module assumes that water is allocated based on land acreage and that farmers pay the same price to withdraw water from its source regardless of their heterogeneities. Through the use of more efficient irrigation technology, LSFs consume only a portion of their water rights and store the remaining share. The water rights stored could then be re-utilized in two ways. First LSFs could sell the remaining water rights to EFs to enable them to complete their irrigation schedules, since EFs have lower irrigation efficiency. Second the remaining water-use rights could be self-consumed by LSFs to expand their irrigation schemes. Expansion of irrigation schemes requires additional irrigation and other input costs (fertilizers, pesticides, etc.). Although farm profit maximization remains the ultimate objective in this model it is important to highlight that additional water management objectives might eventually be targeted in South Africa. Objectives such as equity in water allocation, sustainability of water resources and respect of water use for basic human needs and ecological reserves are also important criteria that may facilitate the reconciliation strategy the country has embarked on, since its democratization in 1994. The paper is organized as follows. Section 2 provides an overview of the status quo of water resources in the Olifants. Section 3 presents the two water policies investigated – compulsory licensing and water market – and discusses their role in water re-allocation objectives. Section 4 presents the methodology developed to simulate impacts of CL and WM on farming activities. The results of the simulation are shown in Section 5. Section 6 concludes the paper.

2. Status of water resources in the Olifants catchment With a population of 3.2 million, the Olifants river basin is located in the northeast of the country and it extends over 54.475 km2 (Fig. 1). The region accounts for 7% of the country’s population and it produces 6% of the GDP [21]. The largest part of that population (almost 67%) is located in rural areas, considered as homelands during the Apartheid regime [22]. The Olifants river basin encompasses different sub-basins and it flows in three different regions (Gauteng, Limpopo and Mpumalanga). The main tributaries of the river are Wilger, Klein Olifants, Elands, Steelport, Moses, Ga-Selati and Blyde [13]. The inflow-stream sources are estimated at 631 mm from rainfall and 1992 million cubic meters (Mm3) for runoff [23]. Rainfall occurs mainly in summer and it varies across the catchment. [24] found for example a coefficient of variation of 0.25 for rainfall. Runoff also varies across the catchment with an 4 A different definition of large-scale and emerging farmers can be found in [6]. Moreover [17–19] categorized EFs (smallholder) farmers in four categories throughout the country (1) farmers in irrigation schemes, (2) independent irrigation farmers, (3) community gardeners and (4) home gardeners. [20] identified approximately between 200 000 and 250 000 smallholder irrigators throughout South Africa.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

4

Fig. 1. The Olifants region, South Africa.

annual average of 37.5 mm accounting for 6% of the annual rainfall [25]. The annual discharge sum of the whole basin is 2040 Mm3. Groundwater abstraction varies between 75 and 100 Mm3 with mines, irrigation and households being the main users of groundwater [26]. The Olifants catchment is stratified into Upper, Middle and Lower Olifants [21,26,27]. Irrigation agriculture is the main water consumer in the basin (Table 1). The main agricultural activities are located around the Loskop dam, the Steelport valley and upper Ga-Selati catchment [21,26]. [28] identified a large variety of farmers in the catchment: (i) small-scale subsistence and (ii) large-scale commercial farmers. As in many parts of South Africa, the water demand has increased in the Olifants due to an increase in water use for agriculture and electricity production, economic development and as a result of population growth [22,27,29]. In order to manage the new water consumption pattern, additional dams were constructed in the basin. [30] identified 37 major,5 300 minor6 and around 4000 small dams7 constructed mainly for irrigation purposes and livestock watering. The increase in water consumption in different sectors has also led to (a) water imbalance between demand and supply in many locations along the catchment and (b) the deterioration of water quality due to the lack of treatment of polluted effluents poured into the river. Table 2 shows the water balance in the upper, middle and lower Olifants. It is shown that water imbalance will persist at least until 2030. For example in 2010 water imbalance was deeper in middle and lower Olifants compared to the upper Olifants. When ecological reserves8 are considered the whole catchment shows a water imbalance in 2030. Beyond water imbalance, the deterioration of water quality is a serious concern in the basin. [31] argues that water pollution is the biggest concern the basin is currently facing. The pollution comes from different sources (i.e., runoff and drainage from irrigation return flow, waste water, discharged mine effluents, seepage, etc.). The deterioration of water quality leads to an increase

5

Major dams are reservoirs storing more than 2 Mm3 volume of water. Minor dams are reservoirs storing between 0.1 and 1 Mm3 volume of water. 7 Small dams are reservoirs storing less than 0.1 Mm3. 8 Ecological Reserve is the volume of water that must be allocated to the environment to preserve the sustainability of the biodiversity. See [33] for more details about ecological reserve in the South African water context. 6

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

5

Table 1 Water consumption in the Olifants (million m3/annum). Source: [26]. Management Zone

Irrigation

Urban

Rural

Industrial

Mining

Power Generation

Total

Upper Olifants Middle Olifants Lower Olifants Total

249 81 156 486

93 56 29 178

4 22 3 29

9 0 0 9

26 28 32 86

228 0 0 228

609 187 220 1016

Table 2 Scenarios of water balance in the Olifants catchment (million m3/annum). Source: [26,27]. Zones/years

2010

2030

2010

2030

Upper Olifants Middle Olifants Lower Olifants Total

618 128 202 948

618 227 202 1047

612 156 218 986

648 214 230 1092

2010

2030

2010

2030

18 18

80 51 69 200

6 (28) (34) (56)

(110) (38) (97) (245)

Fig. 2. Process of Compulsory Licensing. Source: [35].

in treatment costs. Another consequence of water pollution is its eventual impact on health. Diseases such as diarrhea, cholera and other bacterial infections are caused and maintained by bad water quality. Polluted water might also impact on food quality through an increase in microorganisms in crops. Metal accumulation and endocrine disrupting substances present in the water could modify the biological characteristics of crops, making them more difficult to be conserved and compromising their ability to adapt to changing climatic conditions. More recently [32] argue that a clean water source influences food availability, access, utilization and stability, especially in the poorest countries. Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

6

3. Policy of water re-allocation in the Olifants catchment In the NWA, compulsory licensing and water market are seen as important instruments to promote equity, efficiency and sustainability in water allocation. The main objective of these policies was to increase impacts of water on economic growth in South Africa and to restore equity in water allocation mechanisms.

3.1. Compulsory licensing Compulsory licensing is a mechanism introduced in the NWA (Sections 43–48) in [34] to promote a re-allocation of water resources in water stressed catchments in South Africa. The process of implementing compulsory licensing in the agricultural sector follows mainly two steps (Fig. 2). First, the legal authority publishes a notice in the Government Gazette to invite water users in productive sectors (agriculture in our case) to apply for licenses. By submitting their applications, all the water users must declare the quantity of water use to enable a re-assessment of unfair and previous over allocations. Second, the authority issues a proposed allocation schedule. Once that proposed allocation schedule is approved, it is referred to as a preliminary allocation schedule. Curtailments might occur to balance water access discrepancies between users. When a particular user (farmer) raises an objection against the proposed allocation schedule, the responsible authority must consider the claim and integrate it into the preliminary allocation schedule, if the objection is credible.9 When the claimer is still not satisfied with the decision he (she) could bring the case to the Water Tribunal for further inquiry. However, claims are not valid if they go against the two fundamental principles of the National Water Act: (i) against a rectification of over-allocations or an unfair and disproportionate water use and (ii) against the obligation to allocate water to the reserve (ecological reserve and basic human need). The preliminary allocation schedule is followed by the final allocation schedule. Once the final allocation schedule is approved, licenses are issued to replace water entitlements. The licenses give to their holders the rights to use water accordingly. Currently three catchments – Mhlathuze in Kwazulu-Natal, Tosca in North West Province and Jan Dissel in Western Cape – have implemented compulsory licensing.

3.2. Water market Water market is another mechanism incorporated in the National Water Act (1998) to promote a sustainable management of water resources. It allows a voluntary transfer of water use rights,10 in exchange for financial compensation. In the agricultural sector of the Olifants region water market assumes that farms holding licenses that are not used after a completion of irrigation schedules (surplus licenses holders) sell such licenses to the ones that still need additional water (deficit license holders) to complement their irrigation schedules. In such a framework water market ensures a trade of water-use rights though those rights are purchased (sold) without a transfer of land [36,37]. In the country, water reform is seen as an important component of land reform [38–41]. Indeed, in many catchments land claims make farmers hesitant to undertake investments due to uncertainties that might arise after a purchase of water rights. It is even argued that land claims impede water rights’ transfers because of potential risks that might impact the sustainability of farming activities [41]. However, beyond these uncertainties the Government has seen – in water trading as in compulsory licensing – valuable instruments that encourage an efficient re-allocation of water resources across the nineteen water management areas. 9

Objections might only be raised within 60 days following the issue of the proposal allocation schedule. The water use rights are based on a number of licenses allocated by the public authority to farmers, after approval of the final allocation schedule. Therefore it is important to highlight that we do not mean by water use right a sort of riparian-based water right but rather a license-based water right as encountered in the NWA. 10

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

7

4. Model framework We provide an integrated modeling approach combining water and agricultural modules to investigate impact of compulsory licensing and water market on farming activities. The model uses an optimization technique with the objective of profit maximization. The decision variables are quantity of crop produced and water used in irrigation for a given technological and water policy. The model is adapted from [42–44]. To make our model closer to the reality in South Africa, farms are differentiated between large-scale (LSFs) and emerging (EFs) groups with different levels of heterogeneity in their types, e.g. land acreage, technology use for irrigation and past historical heritage. The large-scale farms hold more than 150 ha of land and they use efficient irrigation technology (sprinkler irrigation). Moreover, due to a longer experience in the sector they have better knowledge of farming activities and a better learning ability than the EFs. This learning ability increases the efficiency of irrigation measures and it allows a much more significant reduction of long-term irrigation costs. A better learning ability also minimizes water losses within the water-use-chain since water use per unit of output produced becomes more efficient. We also assume that the LSFs have available acreages we denote latent lands, which could be used later on if the right market and incentive conditions prevail. Contrary to the LSFs, the emerging farms are the previously disadvantaged population with limited land acreages (10 ha), a poor knowledge of farming activities and a highly vulnerable status about price volatility (both crops and water prices). Emerging farms do not hold latent lands for agricultural expansion and their water uses are either from the national regulator (share of water available for allocation) or from water traded (acquired) from large-scale farms. By analyzing the impacts of water policy on farming activity two scenarios are built, which are linked to each of the selected policy instruments. We first assume a curtailment of entitled water rights from LSFs to EFs. Within compulsory licensing curtailment might be an option to re-balance past discriminatory practices and to support the rise of emerging farmers who lack access to water rights. We simulate a curtailment of 20%, 30% and 40% of water rights allocated to LSFs and analyze the impacts of a re-allocation of those shares to EFs.11 Second our model provides economic conditions under which emerging farms invest in more efficient technology to increase water saving and water use efficiency. The analysis also provides economic conditions under which water trade remains beneficial for both emerging and large-scale farms with regards to their respective economic and technical assets. Fig. 3 shows the framework of the model. 4.1. Model structure

MaxΠ n ¼ ðΠ L þ Π E Þ

with n ¼ ½L; E

ð1Þ

subject to wan þ S ¼ R S40 R40

ð2Þ

where Π n denotes the profit function of farms, with Π L and Π E representing profits of representative large-scale (LSFs) and emerging (EFs) farms, respectively. S and R are the instream flow and the available source of water, respectively. Eq. (2) assumes that water allocated to farms and that required in in-stream flow could not exceed the available volume of water in the catchment. wan is the volume of water allocated to farms for irrigation purposes. We assume that such a volume is split in waL and waE representing the volume of water allocated to large-scale and emerging farms respectively. Through a use of more efficient irrigation technology, LSFs consume only a portion of the allocated 11 The underlying assumption is all LSFs are assumed to be equally efficient in water use and equally efficient across all crop types. In real life, one might also expect water trades amongst LSFs either because they are not equally efficient or because they produce different crops, which produce different efficiencies. But this assumption is made to keep the analysis and the interpretation of the results as simple as possible.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

8

Fig. 3. Framework of the model.

water rights wCL ¼ αwaL with 0 ! α ! 1. Therefore the water saved by LSFs wSL ¼ ð1  αÞwaL could be sold to emerging farms, which use less efficient irrigation technology. Hence the water available to emerging farms is wCE ¼ waE þ ð1  αÞwaL , representing the sum up of water allocated by public authority and the remaining portion of water saved by LSFs the EFs bought. The model highlights the possibility for EFs to invest in more efficient irrigation technology, which results in an increase in water use efficiency. Such an increase of water use efficiency reduces the volume of water bought from LSFs. 4.2. Policy scenario of CL The policy of CL is analyzed through curtailments applied to farms. The model simulates a curtailment of 20%, 30% and 40% of the initial allocated water rights to LSF and investigates its impact on a particular EF. The model is solved in maximizing profits of farmers. MaxΠ n ¼ PN n FC n  OC n

ð3Þ

PN n ¼ P i  ϕin

ð4Þ

with

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

FC n ¼ X n þ

1 2 q δ in

9

ð5Þ

OC n ¼ pw1  wan

ð6Þ

X n ¼ ψQ on þ βn wcn

ð7Þ

where ðPNn Þ is the income, ðFC n Þ is the tenure payment for agricultural sector and the tenure payment
 for water sector is ð
OC nÞ. Eq. (4) depicts the evolution of income, which depends on output prices pi and yield of crops ϕin . The model simulates maize, cotton, vegetables and wheat as such grains are among the main crops cultivated in the region (Council for Science and Industrial Research, 2013).12 Therefore, we denote (i) the type of crops considered. Eqs. (5) and (6) are the tenure for the two different modules: agriculture and water. The tenure of the agricultural sector is split in irrigation and other input costs such as fertilizers and pesticides. The irrigation cost ðX n Þ has a fixed ðψQ on Þ and a
 variable βn wcn component, as pointed out in Eq. (7). The fixed component is composed of an initial unit capital cost ðψ Þ and the capacity of the technology ðQ on Þ with ðQ oL g Q oE Þ.13 The variable part of the irrigation cost is a function of water consumed by farms. As mentioned above, beyond irrigation costs, we also consider other input unit costs ðδÞ representing in our example fertilizers used to improve soil quality and yield of crops. We assume that the amount of fertilizers used is proportional to the
 quantity of crops produced qni , with a quadratic relation. The tenure of the water module (Eq. (6)) shows the costs associated with the initial volume of water allocation. We denote pw1 , water price paid by farms to the public authority in their early allocation phase. Such a price might be seen as the payment needed to divert water from its source to the farm. The model highlights that both LSFs and EFs pay the same initial water allocation price. Therefore the cost associated with water delivery is the unit price of a volume of water multiplied by the total volume of water allocated to farmers. Our model introduces irrigation efficiency represented by ðhn Þ to capture the difference in irrigation technology used between the two types of farms. For simplicity reasons, without a loss of generality, we postulate that irrigation efficiency depends on both land quality and type of technology used for irrigation purposes.14 We assume that both farm types have the same land quality but they differ in irrigation technology used. In accordance to [45], we assume a linear relationship between water used by crops ðen Þ and applied water ðwc Þ with en ¼ hn ðX n Þwcn , where hn ðX n Þ represents irrigation efficiency of farms with ðhL g hE Þ.15 The irrigation efficiency is a percentage, ranging between ½0; 1 and it captures the volume of water effectively absorbed by crops after evapotranspiration. Moreover we differentiate between the parameter of irrigation costs of different farms X n  ½X L ; X E , therefore irrigation efficiency is defined by hn ðX n Þ ¼ ðX IM  X INM =X INM Þρn where X IM is the irrigation cost when investment to increase efficiency is made whereas X INM represents the irrigation cost when there is no investment made to improve irrigation efficiency. ρn  0; 1½ with LSFs, since they use more efficient technology whereas ρn ¼ 1 with emerging farmers which use less efficient technology. Following [42]
 we assume that the crop-response function takes a von Liebig form with min hn wcn ; qnin . Such a function assumes that crop growth is tied to a use of a minimum amount of inputs, which are considered as requirements for the plant’s growth. Finally in accordance with [46] we describe the yield of a given crop as a linear function of cumulative evapotranspiration (ET) and assume that wc ¼ qni  a0 An =bn An . The farms maximize profits with respect to output and water used, yielding to following first-order conditions: dΠ n 1 2 ¼ pi hn   q ¼0 bn An δ ni dqni

8 n ¼ L; E

ð8Þ

12 Face to face interview conducted with Dr. Marius Claasen, Manager Water Resources Portfolio, Council for Science and Industrial Research (CSIR), South Africa–March 2013. 13 Because the technology needed to irrigate the acreages of LSFs is higher than the one needed for EFs. 14 Indeed many other factors may impact the irrigation efficiency (type of irrigation system, type of crop and growth stage, irrigation management, environment etc.). 15 Because irrigation efficiency is higher with LSFs compared to the emerging ones, as assumed.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

10

 dΠ n 1
p þ λ1 ¼ pi hn βn  α w1 dwcn dΠ n ¼ pi hn βn  pw1 þ λ1 dwcn

if n ¼ L

if n ¼ E

ð9Þ ð10Þ

with 0 ! λ1 ! 1 Lagrange multiplier. Eqs. (8)–(10) show the evolution of crops and water shadow prices, when the policy of compulsory licensing is considered. The underlying assumption with that policy is that product prices are independent to farm's size and water is allocated per hectare, therefore both large-scale and emerging farmers pay the same water price to the public regulator. The equations also give the upper limits of
 water and output prices pi ; pw1 above which the farms find no more incentives to produce crops (break even point). One can remark that crop production is an increasing function of inputs used (water and fertilizers) and a decreasing function of evapotranspiration. 4.3. Policy scenario of water market We follow [42,47] and consider water market as transactions of water-use rights16 between farms (in this case large-scale and emerging farms). To simplify the model, we assume that only large-scale farmers are entitled to sell a remaining portion of their water rights wsL ¼ ð1  αÞwaL to emerging farmers, since they use more efficient irrigation technologies (high hn and low βn ) than emerging farmers (low hn and high βn ). The model also highlights the possibility for LSFs to make a tradeoff between selling the remainder of their water rights to emerging farmers and using those rights to produce additional crops. We assume that once LSFs use their remaining water rights, crops produced with their latent lands are sold in the agricultural market at the same price as the first crops produced before the settlement of water market. Therefore there is no price incentive to transfer water use to a second phase in order to take a benefit from a potential rise of crop prices. Moreover, the use of latent lands is subject to the payment of additional variable costs (fertilizers and water use). The decision of LSFs is taken after a comparison of outcomes between using the remaining water rights for additional crop production and their sales to emerging farmers. Eqs. (11) and (12) represent the outcomes of these two options where Π RU=L and Π RS=L represent the profits of LSFs when the remaining rights are self-used and sold, respectively. pw2 is the price of the transaction capturing the water price LSFs charge to EFs (water traded price). Π RU=L ¼ pi hL ð1  αÞwaL 

1 2 q  βL ð1 αÞwaL δ iL

Π RS=L ¼ pw2 ð1  αÞwaL

ð11Þ ð12Þ

Therefore the large-scale farms’ decisions to trade water depend on the comparison between the income earned in selling the remaining water rights (water rights stored) and the one earned while using the rights. If G (…) represents the difference between Π RU=L and Π RS=L , the sign of G (…) determines the decision of LSFs. Three possibilities might arise; G (…) 40 when LSFs self-use the remaining water rights stored, G (…)o0 when water market is more beneficial and G (…) ¼0 when LSFs are indifferent between using and selling the remaining rights stored.
 G ¼ v pw2 ¼ Π RU=L Π RS=L ð13Þ Substituting in Eqs. (11)–(13) and simplifying, the decision of LSFs is represented in Eq. (14). Z Z 1

 1 1 2  ηt wsL pw2  k e  ηt dt þ qni e dt ð14Þ V pw2 ¼ δ 0 0
 with k ¼ pi hL þ βL The LSFs’ decision depends on the present return of the increase in income over all future time period. Since variables are deterministic, they are discounted by the risk-free interest rate η. If we pose 16

One may also say water licenses, according to footnote 9.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

11

that lim ex ¼ þ 1 and lim ex ¼ 0, the resolution of Eq. (14) leads to the following condition. x- þ 1

x-  1

1 1 pw2 g k  q2ni  δ wsL

ð15Þ

The result is intuitive since it tells us that LSFs are more likely to trade water with EFs as long as the price of the transaction (water traded price) allows a recovering of both fertilizer and water saved costs. The water saved costs embedded all the storage facilities mobilized to maintain the volume of water saved by LSFs in a tradable condition. It is also important to highlight that Eq. (15) shows that an increase in crop production decreases the likelihood of water trade instead of an increase of water saved which stimulates water trade. Like the LSFs, EFs have two options to increase their water volume for a completion of their irrigation schedules: a purchase of the remaining water rights from large-scale farms wsL ¼ ð1  αÞwaL or an investment in more efficient irrigation technology which results to a use of less water per output produced (an increase of hn ). The decision of the emerging farms with the creation of the water market is twofold. First it compares the expenditures from a purchase of water rights sold by largescale farms SRB=E (Eq. (16)) with the value of water saved due to an investment in new technology W SE (Eq. (17)). The value of water saved measures the opportunity cost of water trade. The main objective of using more efficient irrigation technology is to improve water use efficiency, which leads to an increase in water saved during irrigation processes. In addition, more efficient irrigation technology reduces the deterioration of water quality due to a reduction in runoff and drainage discharged by farmers to the river.17 Second the emerging farmers make a trade-off between the investment made for an improvement in irrigation technology and the irrigation efficiency reached. Eq. (18) shows the relationship between irrigation efficiency and cost levied for investment. An increase in investment in irrigation technology improves irrigation efficiency, which contributes to an increase in the amount of water saved. When the LSFs decide not to sell their remaining water rights, then the only alternative that remains for EFs to complete their irrigation schemes is to invest in more efficient technology to increase water saving. Fig. 4 shows the problem tree analysis of the water market. SRB=E ¼ pw2 ð1  αÞwaL

ð16Þ

W SE ¼ waE wcE

ð17Þ

 hn ðX n Þ ¼

X IM  X INM X INM

ρn

ð18Þ

with X INM ¼ ψQ on þ βn wcn 0

X IM ¼ ψ 0 Q on þ βn wcn

ð19Þ ð20Þ

In combining Eqs. (15)–(20) one can write Eq. (21) which links investment in irrigation technology with irrigation efficiency. Water saved is the difference between water allocated and water consumed. 0

hn ¼

ψ 0 Q on þβn wcn 1 ψQ on þ βn wcn

ð21Þ

After a combination of Eqs. (16)–(21), the volume of water saved could be written as a function of irrigation efficiency, in Eq. (22). wsE ¼ waE  with

Δψ ¼ ψ 0 ψ

17

Q 0on ðhn ψ  Δψ Þ Δβn  hn βn

and

ð22Þ

Δβn ¼ β0n βn

Moreover, higher irrigation efficiency lowers groundwater recharge and this reduces inputs into the drainage system.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

12

Fig. 4. Problem tree of water trade.

4.4. Model calibration The model has been calibrated using a compilation of different datasets from various South African water and agricultural research sources (WRC, DWA, ARC, SAPWAT).18 Those datasets are completed by additional data from a farm-survey undertaken by [48] during the first phase of the Olifants project financed by the Federal Ministry of Education and Research (BMBF) of Germany. Three different data sources are needed to run the model; water, agricultural module parameters and techno-economic data. The model is calibrated for the year 2007 in order to match the characteristics of the different data sources. The parameters of the water module are taken from different reports published by the above mentioned institutions. The knowledge hub of the Water Research Commission of the Republic of South Africa has allowed a compilation of a large range of data in the region. Therefore we scrutinized more than 45 research reports ranging from different parts of the country and drawn average data about the water parameters in the Olifants. We simulate the model using the data within the 95th percentile to minimize the variations that might occur due to the large variability in the data sources. Those data are later on compared with some statistics drawn from SAPWAT, a program developed to estimate water requirements of crops and farms in South Africa. Moreover, water allocated, irrigation costs and volume of water used by crops are taken from [48]. The agricultural module parameters are taken from the Agricultural Research Council and literature [49–51]. To consider the time lag, price data are computed considering the inflation of the country between the year 2000 and 2007. Crop types and farms’ behaviors in the region are also specifically depicted in this analysis.

5. Results The following example illustrates the effect of policy of the compulsory licensing and water market on crop production, water used and investment decision in irrigation technologies. The model first 18

WRC: Water Research Commission, DWA: Department of Water Affairs, ARC: Agricultural Research Council.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

13

considers the baseline scenario in which no curtailment is introduced. When no curtailment is introduced both LSFs and EFs produce outputs based on the evolution of crop prices, irrigation efficiency, land acreages, price paid to withdraw water from its source and water price paid when water market is introduced. As expected, curtailment reduces both crop production and income of large-scale farms, though the scope of the reduction differs from crop to crop. Our results show for instance that LSFs reduce more the production of vegetables and cotton than wheat and maize. Indeed with a curtailment of 40% of the initial allocated water, production of wheat and maize are reduced by 0.68 t/ha and 0.64 t/ha whereas the production of cotton and vegetables are reduced by 0.8 t/ha and 0.72 t/ha, respectively. With a curtailment of 40% the profits of maize and wheat are reduced by $32/ha and $30.12/ha whereas the profits of cotton and vegetables are reduced by $35.6/ha and $46.12/ha, respectively. The production with wheat and maize was less influenced by curtailments compared to vegetables and cotton because farmers can easily find a secure and a sustainable market for wheat and maize crops compared to other crops. Wheat and maize represent very important grain crops in the South African agricultural portfolio. For instance the Department of Agriculture, Forestry and Fisheries (2012) ranks wheat and maize grains among the third most important crops produced in the country. On the other hand, curtailments impact less on wheat production because such crops are less dependent on water allocated by public authority, for their growth. Indeed in the country wheat is usually cultivated in winter and summer rainfall seasons (between mid-April and mid-June and mid May and end of July). LSFs also reduce their maize production less than vegetables and cotton because maize prices have increased during the period considered leading to an increase of income to farmers. Indeed, maize is also one of the most stable crops in the agricultural portfolio of the country [52,53]. The crop is used for human (white maize) and animal (yellow maize) consumption (Table 3). For the emerging farmers curtailment is also found to increase both crop production and income, as expected. However it is important to highlight that the reduction in crops and income of LSFs associated with curtailments of water rights is not fully transferred to EFs’ earnings, since the sum of revenue and crop production lost by LSFs is not fully equal to the sum of revenue and additional crops earned by EFs. For example with a curtailment of 30% LSFs reduce their incomes by $20.8/ha; $27.32/ha, $26.92/ha and $17.44/ha for maize, cotton, vegetables and wheat respectively whereas EFs Table 3 Baseline scenario. Water module parameters waE Initial water allocated to emerging farmers (EFs) waL Initial water allocated to large-scale farmers (LSFs) ψ Average initial unit capital cost 0 Average unit capital cost, after investment ψ βn Average initial variable irrigation cost βn' Average variable irrigation cost, after investment pw1 Price of water allocated Irrigation efficiency hn en Average water use by crops pw2 Price of water traded α Share of water allocated used Agricultural module parameters Reference evapotranspiration an Crop evapotranspiration bn Acreages An pi Crop prices Cotton Wheat Maize Vegetables Techno-economic parameters ρi Elasticity of effective water

8. 000 m3/ha 8. 000 m3/ha 12054 R/ha 15.000 R/ha 10 515 R/ha 11 320 R/ha 3.7 c/m3 85% (LSFs) and 50 % (EFs) 350 mm 14.3 c/m3 0.4 680 mm 538 mm 150 ha; 10 ha R R R R

2.65/kg 1.150/t 900/t 1.200/t

0.8 for EFs ¼1

Data compiled from different sources: SAPWAT, WRC and Linz (2010).

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

14

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

Fig. 5. Change (decrease) in profits with curtailments in water-use rights: Large-scale farmers (LSFs).

Fig. 6. Change (decrease) in production with curtailments in water-use rights: Large-scale farmers (LSFs).

increase their incomes by $18/ha, $6.4/ha, $11.2/ha and $16.8/ha for the respective crops. With a curtailment of 40% LSFs see their income reduced by $25.2/ha, $35.56/ha, $46.12/ha and $30.12/ha for maize, cotton, vegetables and wheat respectively while the emerging farmers will have an increase of $21.2 /ha, $13.72/ha, $14/ha and $23.6/ha for the respective crops. Therefore one can remark that there are other factors, besides water input, which have a strong impact on farmers’ outcomes. Figs. 5–8 show the evolution of benefits and losses drawn from the curtailment scenarios. Fig. 9 shows the impacts of investment on irrigation efficiency. It denotes that irrigation efficiency increases with investment made by farmers. Generally such an investment takes two forms. First it can be assimilated to investment made for an acquisition of new irrigation equipment, which are more efficient in water use. An efficient water use occurs when the volume of water withdrawn from the source reached crops with minimum losses during conveyance and the plant can make best use of that water abstracted [50]. For example when emerging farmers use furrow irrigation, an investment made in sprinkler should lead to an improvement in water use efficiency [51]. Second, the investment to improve water use efficiency could also be undertaken in improving irrigation practices. Therefore instead of introducing new irrigation equipment, farmers could discover new techniques and water use practices which increase crop yield while reducing water losses during irrigation. Fig. 10 shows the impact of improvements in irrigation practices on irrigation efficiency. Introducing new technology increases irrigation efficiency more than discovering new practices. The increase in irrigation efficiency will determine the scale with which emerging farmers engage in water market, as represented in Eq. (22). For example an increase in irrigation efficiency increases the volume of water saved by EFs, which reduces the volume of water bought from LSFs. By analyzing the impact of water price charged to EFs (water traded price), Fig. 11 provides the area in which farms engage in water market. Water market occurs if two essential conditions are met. First LSFs should decide to sell their remaining water rights and second EFs should find more benefits to purchasing such rights instead of improving irrigation efficiency. As we assumed that only large-scale farmers are entitled to sell water rights, then variables such as volume of water saved by LSFs wsL , water traded price pw2 and volume of water saved by EFs wsE are the direct factors influencing farmers’ decisions to trade water, once we assume that crop prices are known with certainty. Fig. 11 shows the evolution of water saved in function of water traded price. We note pmin w2 the minimum threshold price required for any water market to take place, as evidenced in Eq. (14). Under the baseline scenario we found a threshold price of 14.3 c/m3. Therefore the area in which water trade is endeavored represents AGFPw2, which we define as area of potential water trade. That area represents Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

15

Fig. 7. Change (increase) in profits with curtailments in water-use rights: Emerging farmers (EFs).

Fig. 8. Change (increase) in crop production with curtailments in water-use rights: Emerging farmers (EFs).

. . . . . . . . .

Fig. 9. Linkages between irrigation efficiency and investment in irrigation technology.

Fig. 10. Linkages between irrigation practices and irrigation efficiency.

the existing water trade possibilities if water traded price pw2 is considered as the only determinant of trade (the lower limit of water traded price). However in introducing the water saved function of the farmers we determine the area of effective water trade (ABPw2). The effective water trade is the difference between the potential water trade and the water saved capacity. Since only the volume of water saved by LSFs is tradable the water saved function represents therefore the border of the water

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

16

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

Fig. 11. Areas of water trade between LSFs and EFs.

trade possibility functions. The set up of effective water trade is undertaken while assuming that EFs have not increased their irrigation efficiency. However in our model it is indeed possible for EFs to increase their irrigation efficiency, which leads to an increase in their volume of water saved. Therefore, the area of effective water trade is conditioned by the evolution of water saved functions wsE and wsL . Under optimal conditions, effective water trade becomes ABC since both EFs as well as opt LSFs use their optimal volume of water saved wopt sE and wsL . 5.1. Policy implications Our results show the importance of putting in place adequate water policies to promote efficient re-allocation of water resources for productive purposes in the Olifants. Through a simulation model, our analysis has shown the potential impacts of compulsory licensing and water market on crop production,19 water used and investment decision in more efficient technology. Compulsory licensing is assessed through curtailments in water allocated. Water market assumes tradable water rights between LSFs and EFs. Although curtailment might allow a re-balance of past discriminatory water allocation practices, care should be given to the scale of that curtailment rate in order to match water re-allocation objectives with agricultural efficiency requirements. Our evidence shows that the losses recorded with LSFs with curtailment are not fully transferred to the EFs. Indeed additional incentives should be provided to EFs in order to improve their irrigation efficiency and their access to the domestic and international agricultural markets. Crop price evolution also plays a key role on farms’ decisions of water use. We also provided a threshold water traded price (14.3 c/m3), which guarantees the settlement of water market. Beyond water traded price we have also shown the impact of investment in irrigation technology on irrigation efficiency. We found that an investment in new irrigation equipment as well as an improvement in irrigation practices increase irrigation efficiency. In terms of water policy our results have at least two policy implications. (1) Both compulsory licensing and water market offer new mechanisms that could be used by public authorities to reshape water allocation practices in South Africa and to promote the rise of EFs. Indeed, by comparing the threshold of water traded price with the unit of investment made per 19 However, it is important to highlight that the levels of incomes losses (gains), cropping patterns and irrigation efficiency reached in our analysis depend on a number of assumptions made throughout the paper. Changes of these assumptions may lead to slightly different outcomes. This will not change, however, significantly our main conclusion, since additional studies have concurred with our conclusions, even after having used a different methodological approach, scale and having introduced more flexibility in the assumptions with regards to land quality, irrigation efficiency and characteristics of water market (Walter et al, 2011).

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

17

unit of irrigation efficiency reached, EFs make optimal decisions about the additional volume of water bought from LSFs. (2) Beyond compulsory licensing and water market, additional incentives are necessary to improve water use efficiency and to connect the water and the agricultural markets. Because the decisions made in the water market (for example using new irrigation technology, buying additional volume of water, changing irrigation practices and introducing volumetric water tariffing) are directly influenced by the agricultural market (crop prices and their potential evolutions, access to market and market structures of the crops).

6. Discussions and conclusion of the study The paper analyzed how compulsory licensing and water market could contribute at reshaping the past discriminatory practices of water allocation between farmers in South Africa. We found that both instruments provide interesting solutions to increasing water access for EFs. Our findings show that the rise of emerging farmers should be comprehended in a broader context, since other factors besides water policy also influence total outcomes. Therefore the water reforms in South Africa should be embedded in a broader context through a combination of water re-allocation instruments with agricultural development programs. Moreover in addition to the combination of water and agricultural development programs, further institutional changes are required in the country to reinforce the impacts water policy may have on agricultural productivity. These institutional changes may take at least two forms: well-defined property rights and an establishment of administrative and bureaucratic systems that reduce transaction costs of implementing water policy such as water market. Different studies have acknowledged the positive effects of well-defined water rights on water market [14,15,42,47,54–57], since this protects ownership and secures long-term investments in the water sector. When property rights are reliable, enforceable and transferable, this reduces risks and uncertainties inherent from the functioning of water market. In South Africa [15] argues that establishment of water market may follow three steps. First water market of irrigation water use rights may be authorized only for farms located within the same WUAs and, in the second step, water market may be extended between farmers and urban users of the same WUAs. Finally, as a third step, water market may then be generalized between farmers located across different WUAs, thereby favoring inter-catchment water transfers. This inter-catchment water trade allows for risk minimization since farms may still acquire water, even when random climatic calamities such as drought occur. In the Olifants river basin, some water-related problems – such as water imbalance between demand and supply, water scarcity and poor water quality standards – have been identified [21,22,27,31]. Many of these problems can be addressed in designing and promoting sound public policies that target both efficiency in water allocation and use and equity in water access. These public policies may contribute to ensuring the ecological reserve and the basic human need, as outlined in the National Water Act (1998). They may also contribute to improving water quality standards across the river and at balancing demand and supply of water in such a way that water resource can be consumed wherever it is needed along the river regardless of the geographical location of farms. This means even farmers located in areas experiencing water scarcity or poor water quality standards may have access to clean water, through water trade. Therefore it is important to promote sound water policies that enhance water allocation in both within and between catchment management areas. The work presented in this paper extends this general assertion in providing a quantitative assessment of impacts of water market and policy of compulsory licensing on farm incomes, crop yield, irrigation efficiency and investment in more efficient irrigation technology. However it is important to highlight that in reality the implementation of such water markets and policy of compulsory licensing is more difficult than usually expected. For instance different experiences across the United States of America and Australia have shown the need to create an enabling environment that incentivizes farmers to trade water and to invest in more efficient irrigation systems [54,55,57–62]. In South Africa, [15] argues that sleeper rights (unused water) may play a good role in the design of water market. In a study applied to the Northern Kwazulu-Natal and in the Lower Orange river, the authors showed that farmers prefer to trade sleeper rights instead of the volume of water saved. Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

18

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

An additional advantage of water market is it allows for efficient pricing of water resources. In South Africa water use for irrigation is subsidized, that means farmers are not paying the full cost that is necessary to ensure viability of the water supply system and to maintain water supply infrastructure. When water is underpriced, it usually leads to over-utilization of the resource, since the price paid for water consumption is lower than the marginal price. This not only has implications on the allocation of the water resource, but it also decreases the water quality standard.

Acknowledgment This research is part of the Integrated Water Resources Management Middle Olifants South Africa (Phase II) Project and is funded by the German Federal Ministry of Education and Research (BMBF). The authors would like to Thank Theresa Link for having provided the data of the farm-survey collected in South Africa during the first phase of the Project. The authors would also like to thank the two anonymous reviewers for their very useful comments and suggestions that have contributed to improve significantly the quality of the paper. The authors would like to thank Marc Müller and Bernhard Tischein for their comments. The authors remain responsible of the remaining errors. References [1] T. Walter, J. Kloos, D.W. Tsegai, Options for improving water use efficiency under worsening scarcity: evidence from the Middle Olifants sub-basin in South Africa, Water SA 37 (2011) 357–370. [2] R. Hassan, J. Thurlow, Macro-micro feedback links of water management in South Africa: CGE analyses of selected regimes, Agric. Econ. 42 (2011) 235–247. [3] G.-M. Lange, E. Mungatana, R. Hassan, Water accounting for the Orange river Basin: an economic perspective on managing a transboundary resource, Ecol. Econ. 61 (2007) (2007) 660–670. [4] R. Hassan, B. Olbrich, J. Crafford, Measuring total economic benefits from water in plantation forestry: application of quasi I–O framework to the Crocodile catchment in South Africa, South. Afr. For. J. 193 (2002) 5–14. [5] D.J. Merry, H. Levite, B. van Koppen, Are Good Intentions Leading to Good Outcomes? Continuities in Social, Economic and Hydro-political Trajectories in the Olifants River Basin, South Africa in River Basin Trajectories: Societies, Environments and Development in Molle, Francois; Wester, P. (Eds.). River Basin Trajectories: Societies, Environments and Development. Wallingford, UK: CABI; Colombo, Sri Lanka: International Water Management Institute (IWMI). Comprehensive Assessment of Water Management in Agriculture Series 8, 2009. [6] T. Walter, Analysis of Sectoral Water Demands and Optimal Water Allocation in the Middle Olifants Sub-Basin of South Africa: The Case of Mining and Irrigation. Ph.D Dissertation, University of Hohenheim, 2010. [7] T. Zhu, C. Ringler, Climate change impacts on water availability and use in the Limpopo river basin, Water 4 (2012) 63–84. [8] R.M. Hassan, J. Crafford, Environmental and economic accounts for water in South Africa, The Economics of Water Management in Southern Africa, 4, Edward Elgar pp.114–166. [9] A. Brown, A Review of Water Scarcity Indices and Methodologies. The Sustainability Consortium, White paper 106, 2011. [10] R. Hassan, J. Thurlow, T. Roe, X. Diao, S. Chumi, Y. Tsur, Macro-Micro Feedback Links of Water Management in South Africa: CGE Analyses of Selected Policy Regimes. Policy Research Working Paper No. 4768. World Bank, 2008. [11] R. Hassan, S. Farolfi, Water value, resource rent recovery and economic welfare cost of environmental protection: a watersector model for the Steelpoort sub-basin in South Africa, Water SA 31 (2005) 9–16. [12] D.B. Louw, The Development of a Methodology to Determine the True Value of Water and the Impact of a Potential Water Market on the Efficient Utilization of Water in the Berg River Basin. WRC Report No 943/1/02, 2001. [13] DWA, A Draft Position Paper for Water Allocation Reform in South Africa: Towards a Framework for Water Allocation Planning. Department of Water Affairs, Pretoria, 2005. [14] R.Q. Grafton, C. Landry, G.D. Libecap, S. McGlennon, Robert O’Brien, Integrated Assessment of Water Markets: Australia, Chile, China, South Africa and the USA. NBER Working Paper No. 16203, 2010. [15] W.L. Nieuwoudt, M. Armitage, Water Market Transfers in South Africa: Two Case Studies in Water Resources Research, 40, 2004, http://dx.doi.org/10.1029/2003WR002840. [16] A. Pott, K. Versfeld, M. van Rooyen, A. Muir, A Scoping Exercise to Investigate the Potential Need for, and Nature of, Water Trading in South Africa. WRC Report No. KV 222/09. South African Water Research Commission, 2009. [17] C.T. Crosby, M. De Lange, C.M. Stimie, I. Van der Stoep, A Review of Planning and Design Procedures Applicable to Smallscale Farmer Irrigation Projects. WRC Report 578/2/00. Water Research Commission, Pretoria, South Africa, 2000. [18] F.J. Du Plessis, I. Van der Stoep, W. Van Averbeke, Microirrigation for Smallholders: Guidelines for Funders, Planners, Designers and Support Staff in South Africa. WRC Report TT 164/01. Water Research Commission Pretoria, South Africa, 2002. [19] W. van Averbeke, Best Management Practices for Small-scale Subsistence Farming on Selected Irrigation Schemes and Surrounding Areas Through Participatory Adaptive Research in Limpopo Province. WRC report TT 344/08. Water Research Commission, Pretoria, South Africa, 2008. [20] G.-R. Backeberg, Reform of User Charges, Market Pricing and Management of Water: Problem or Opportunity for Irrigated Agriculture Irrigation and Drainage 55, 2006, 1–12.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

19

[21] DWA, Classification of Significant Water Resources in the Olifants Water Management Area. Department of Water Affairs, Pretoria. 2011. [22] International Water Management Institute (IWMI), Baseline Report Olifants River Basin in South Africa. IWMI, Sri Lanka, 2008. [23] N. Post-Uiterweer, et al., Redressing inequities through domestic water supply: a ‘poor’ example from Sekhukhune, South Africa, in: Sylvain Perret, Stefano Farolfi, Rashid Hassan (Eds.), Water Governance for Sustainable Development: Approaches and Lessons from Developing and Transitional Economies, Earthscan, London 2006, pp. 54–74. [24] M.P. McCartney, D.K. Yawson, T.F. Magagula, J. Seshoka, Hydrology and Water Resources Development in the Olifants River Catchment. Working Paper 76. Colombo, Sri Lanka: International Water Management Institute (IWMI), 2004. [25] M.P. McCartney, R. Arranz, Evaluation of Historic, Current and Future Water Demand in the Olifants River Catchment, South Africa. Research Report 118. International Water Management Institute (IWMI), 2007. [26] DWAs, Development of a Reconciliation Strategy for the Olifants River Water Supply System, Department of Water Affairs and Forestry, Pretoria, 2011. [27] DWAs, Development of a Reconciliation Strategy for the Olifants River Water Supply System. Department of Water Affairs, Pretoria, 2010. [28] Water Research Commission, WRC Project No 1806: Technology Transfer on the Technical Aspects and Cost-estimating Procedures of Surface and Sub-surface Drip-irrigation Systems. Knowledge for Growth and Development: Knowledge Review 2007/08. Water Research Commission, Pretoria, 2008. [29] B. Van Koppen, Redressing inequalities of the past from a historical perspective: the case of the Olifants basin, South Africa, Water SA 34 (2008) 432–438. [30] Department of Water Affairs and Forestry, Proposal for the Establishment of a Catchment Management Agency for the Olifants Water Management Area, 2002. [31] Council for Scientific and Industrial Research (CSIR), A CSIR Perspective on Water in South Africa. CSIR, Pretoria, 2010. [32] T. Wheeler, J. von Braun, Climate change impacts on global food security in science, 341, 2013, http://dx.doi.org/10.1126/ science.1239402. [33] K. Leenderste, S. Mitchell, J. Harlin, IWRM and the environment: a view on their interaction and examples where IWRM led to better environmental management in developing countries, Water SA 34 (2008) 691–698. [34] Republic of South Africa (RoSA),. National Water Act, Act 36 of 1998. [35] Q. Quibell, et al., Transforming legal access to water to redress social inequality and economic efficiency in Transforming water management, in: Barbara Schreiner, Rashid Hassan (Eds.), Springer, South Africa, 2011. [36] A. Dinar, M.W. Rosegrant, R. Meinzen-Dick, Water Allocation Mechanisms: Principles and Examples, Policy Research Working Paper Series 1779, The World Bank, 1997. [37] G.G. Gillit, W.L. Nieuwoudt, G. Backeberg, Water market in the lower Orange river catchment in South Africa, Agrekom 44 (2005) 363–382. [38] P. Woodhouse, Reforming land and water rights in South Africa, Dev. Change 43 (2012) 847–868. [39] H. Funke, I. Jacobs, Integration Challenges of Water and Land Reform: A Critical Review of South Africa, in Current Issues of Water Management ed by Uli Uhlig, Intech, 2010. [40] B. Van Koppen, H. Sally, M. Aliber, B. Cousins, B. Tapela, Water Resources Management, Rural Redress and Agrarian Reform. Development Planning Division. Working Paper Series No.7, DBSA: Midrand, 2009. [41] M.L. Msibi, P.Z. Dlamini, Water Allocation Reform in South Africa: History, Processes and Prospects for Future Implementation. WRC Report No. 1855/1/11, 2011. [42] J.M. Carey, D. Zilberman, A model of investment under uncertainty: modern irrigation technology and emerging markets in water, Am. J. Agric. Econ. 84 (2002) 171–183. [43] M. Caswell, E. Lichtenberg, D. Zilberman, The effects of pricing policies on water conservation and drainage, Am. J. Agric. Econ. 72 (1990) 883–890. [44] D.-R. Thiam, Policy instruments for a market penetration of low carbon technologies in developing nations, Int. J. Energy Sector Manage. 6 (2012) 465–487. [45] C. Dridi, M. Khanna, Irrigation technology adoption and gains from water trading under asymetric information, Am. J. Agric. Econ. 87 (2005) 289–301. [46] W.Z. Liu, et al., Interrelations of yield, evapotranspiration and water use efficiency from marginal analysis of water production functions, Agric. Water Manage. 56 (2002) 143–151. [47] Y. Tsur, A. Dinar, The relative efficiency and implementation costs of alternative methods for pricing irrigation water, World Bank Econ. Rev. 11 (1997) 243–262. [48] T. Linz, Analysis of Sectoral Water Demands and Optimal Water Allocation in the Middle Olifants Sub-Basin of South Africa: The Case of Mining and Irrigation, Ph.D Dissertation, University of Hohenheim, 2010. [49] R. Tren M. Schur, Olifants River Irrigation Schemes. Report, International Water Management Institute (IWMI), Colombo, Sri Lanka, 2000. [50] F.B. Reinders, et al., Standards and Guidelines for Improved Efficiency of Irrigation Water Use from Dam Wall Release to Root Zone Application: Guidelines. WRC Report No. TT 466/10. Water Research Commission, Pretoria, South Africa. 209 pp, 2010. [51] M. Fanadzo, C. Chiduza, P.N.S. Mnkeni, Overview of smallholder irrigation schemes in South Africa: relationship between farmer crop management practices and performance, Afr. J. Agric. Res. 5 (2010) 3514–3523. [52] Department of Agriculture, Forestry and Fisheries, (DAFF), Wheat Production Guideline, DAFF, Pretoria, South Africa, 2012. [53] Department of Agriculture, Forestry and Fisheries, (DAFF), Maize Production Guideline, DAFF, Pretoria, South Africa, 2012. [54] R. Howitt, Spot prices, option prices, and water markets: an analysis of emerging markets in California, in: K. Easter, M. Rosegrant, A. Dinar (Eds.), Markets for Water: Potential and Performance, Kluwer Academic Publishers, Boston 1998, pp. 119–140. [55] M. Rosen, R. Sexton, Irrigation districts and water markets: an application of cooperative decision-making theory, Land Econ. 69 (1993) 39–53. [56] R. Young, Why are there so few transactions among water users? Am. J. Agric. Econ. 68 (1986) 1143–1151.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001

20

D.R. Thiam et al. / Water Resources and Economics ] (]]]]) ]]]–]]]

[57] Z. Donohew, Property rights and Western United States water markets, Aust. J. Agric. Resour. Econ. 53 (2009) 85–103. [58] C. Howe, J. Lozo, K. Weber, The economic impacts of agriculture to urban water transfers on the area of origin: a case study of the Arkansas River Valley in Colorado, Am. J. Agric. Econ. 72 (1990) 1200–1204. [59] H. Bjornlund, P. Rossini, An Empirical Analysis of Factors Driving Outcomes in Markets for Permanent Water—An Australian Case Study, Centre for Land Economics and Real Estate Research, School of International Business, University of South Australia, 2006. [60] National Water Commission, Australian Water Resources 2005: A Baseline Assessment of Water Resources for the National Water Commission, Key Findings of the Level 2 Assessment: Summary Results, Commonwealth of Australia, Canberra, (2007). [61] V. Wayne, C. Son, Regulating the Australian water market, J. Environ. Law 22 (2010) 431–459. [62] R.-Q. Grafton, J. Horne, Water markets in the Murray–Darling Basin, Agric. Water Manage. 145 (2014) 61–71.

Please cite this article as: D.R. Thiam, et al., Implications of water policy reforms for agricultural productivity in South Africa: Scenario analysis based on the Olifants river basin, Water Resources and Economics (2014), http://dx.doi.org/10.1016/j.wre.2014.11.001