REPUBLIC OF BOTSWANA BOTSWANA S NATURAL RESOURCE ACCOUNTS: WATER ACCOUNTS: PHASE 1. Prepared by CSO AND NCSA GOVERNMENT OF BOTSWANA

Transcription

1 REPUBLIC OF BOTSWANA BOTSWANA S NATURAL RESOURCE ACCOUNTS: WATER ACCOUNTS: PHASE 1 Prepared by CSO AND NCSA GOVERNMENT OF BOTSWANA DECEMBER 2001

2 TABLE OF CONTENTS Acknowledgement EXECUTIVE SUMMARY 1. BACKGROUND 2. FRAMEWORK FOR WATER ACCOUNTS AND SUMMARY OF DATA SOURCES 2.1. Type of accounts 2.2. Classification of water 2.3. Data sources and major data problems 3. OVERVIEW OF WATER USE IN BOTSWANA (1993 AND 1998) 3.1. National Trends in Water Use 3.2. Water Use by Economic Sector 3.3. Supply Losses 4. ECONOMIC BENEFITS OF WATER USE BY SECTOR 5. INTERNATIONAL COMPARISONS OF WATER USE 6. COSTS AND REVENUES 6.1. Water Utilities Corporation 6.2. Department of Water Affairs 6.3. District Councils and self-providers 7. CONCLUSIONS AND THE WAY FORWARD APPENDIX A: Data Sources APPENDIX B: Detailed Water Accounts Data

3 Acknowledgement This report was prepared by the National Conservation Strategy Agency and the Central Statistics Office with the assistance of Dr. Glenn-Marie Lange of the Institute for Economic Analysis, New York University, United States of America. The assistance was funded under the USAID Co-operative Agreement No. 890-A , that established the Southern African Natural Resources Accounting (NRASA) project for Botswana, Namibia and South Africa. NRASA was programmed to assist and enhance the in-country capacity of the three countries to prepare and utilise Natural Resource Accounts in the decision-making process to achieve sustainable natural resources development. The following key personnel, the late Mr Marshall Moseki, Ms. Sarah Kabaija, Ms. Minkie Pheto and Mr James Katale all of the Central Statistics Office (Ministry of Finance and Development Planning), Ms. Mokgadi Monamati and Mr. Edward Karkari of the National Conservation Strategy (Co-ordinating) Agency (Ministry of Lands, Housing and Environment) formed the Technical team to construct the accounts. The Technical team was aided by Professor Jaap Arntzen of the Centre for Applied Research, Gaborone, Botswana. The construction of the accounts also involved members of staff from the Department of Water Affairs, Water Utilities Corporation and from the Department of Local Government Services who represented the institutions as core teams in the collation of the respective primary data sets. The interest shown and support given by the Heads of these institutions contributed largely to the success achieved in the Phase 1 of this exercisee. Such interest and support have been found to be invaluable and it is anticipated that the support especially would be maintained first and foremost in improving on data compilation and storage within each institution to sustain future work (Phase 2) in up-dating the accounts. The primary aim is to enhance decision-making in the wise use and development of our water resources to sustain national development. National Conservation Strategy Agency MLHE.

4 1. OVERALL TREND BOTSWANA S NATURAL RESOURCE ACCOUNTS: WATER ACCOUNTS (PHASE 1) EXECUTIVE SUMMARY Total water use in Botswana has increased only 5% over the 5 years from 1993 to 1998, from about 143 to 150 million cubic meters although the economy and population have grown much faster. (Economic growth rate and the population growth rate over the 5 year period could be appropriate points of references to emphasis the significance of the increase of the total water use The economy relies increasingly on formal water supplying institutions (i.e. WUC, DWA and DC) instead of self-providers. WUC, DWA and DC supplied 45% of water used in 1998, up from 39% in Botswana relies on groundwater for most of its water needs (60%), but dams are slowly contributing a larger share (Table 1, Figure 3). 2. WATER USE BY ECONOMIC SECTORS The Agricultural sector is the major user of water in most countries and Botswana is no exception to this global trend. A significant proportion of 50% of the total water use in 1993 was for agriculture, comprising of about 35% in the traditional sector, dominated by the livestock industry, and the other 15% in commercial agriculture which is mostly irrigated crop farming. (Table 2). The use of water by the Government sector increased by 30%, thereby increasing its share of total water use in the economy from 5% to 6.6% between 1993 and Following agriculture, households constitute the second largest user of water, and one that is rising faster than most sectors of the economy. Though total water use increased by only 5%, household water use increased by 21%, increasing its share from 28% in 1993 to 33% in UNACCOUNTED WATER AND LOSSES WUC undertook a programme to reduce physical leakages in the city system. The result was a dramatic reduction in water losses, from 30% to 14% in 1998 and to 9% in The average rate of unaccounted-for water across all 17 villages has declined somewhat between 1993 and 1998, from 28% in 1993 to 24% in However, there is a tremendous range of loss rates for the different villages supplied by DWA. In 1993 the lowest reported loss rate was 6% for Letlhakane and the highest was 49% in Mogoditshane.

5 District Councils and self-providers together account for a majority of water use in the country but do not provide estimates of losses. If we assume that losses of other providers are similar, the country s total loss rate is 15.7%. 4. COMPARATIVE ECONOMIC CONTRIBUTION BY SECTOR RELATIVE TO WATER USE Agriculture, using 50% of all water in 1993, contributed only 4% of national income (valueadded) and 2% of formal sector employment, if full-time farmers were considered part of the workforce, agriculture's share of employment increases to 32%. Within agriculture, the traditional sector makes a greater economic contribution, relative to the share of water it uses, than the commercial sector, mainly because commercial sector water use is dominated by irrigation, which requires much higher amounts of water for the income generated than does livestock. Mining makes a strong economic contribution in terms of income (35%), but not employment (under 4%), relative to its water use (10%). Manufacturing shows considerable variation the water-intensive food and beverage sectors generate less income and employment relative to the water they require compared to other manufacturing sectors. The Service sectors also show variation in water intensity and economic contribution. The economy is producing higher levels of GDP for a given amount of water used. GDP per cubic meter of water used rose from nearly P 77 to P 96, an increase of 26%. This increase may have been achieved in two ways: 1)-water conservation measures or, 2) through a change in the structure of the economy. Because water efficiency is an important goal, more work should be undertaken to determine the reason for increasing macroeconomic water efficiency. In Agriculture, the economic contribution of water per cubic meter of water use declined between 1993 and 1998, from P 6.75 to P This change resulted mainly from the decline in livestock, a more valuable use of water than irrigation (Table 2). The economic efficiency of water use in Mining increased by an impressive 25%, from P 244 per cubic meter of water use in 1993 to P 302 in 1998, largely due to water conservation measures in diamond mining. The economic efficiency of water use in the Service sectors also rose in this instance by 51% from P 927 in 1993 to P 1404 in INTERNATIONAL COMPARISONS Comparing three countries in the region, without prejudice to the differences in socioeconomic settings and water resource endowments for Botswana, Namibia, and South Africa, Botswana has the lowest per capita use of water at 95 cubic meters per person per year in Namibia has a per capita consumption of 144 cubic meters of water per person whilst South Africa s per capita consumption stood at 412 cubic meter per person. Adjusting for agriculture, Botswana and Namibia were very similar, though South Africa still consumed much more water per person.

6 The average Pula of GDP per cubic meter of water was P 124 for Botswana, P 45 for Namibia, and only P20 for South Africa. 6. COSTS AND REVENUES Very little progress has been made in this aspect of the study so far because information about costs of supplying water and the basis of revenue is often not available from water suppliers or is very incomplete. Review of information from WUC shows that all the existing tariff rates increased between 1993 and 1998, and the increase was greatest for the rates for standpipes. At the time of the review, information was not accessible about the costs of supplying water to different towns. With the review of information at the DWA serious gaps in the cost and revenue data were evident and for instances several months' data were found missing out of any particular year. It was therefore impossible to estimate the missing figures for costs and revenues. No cost and revenue data were available from District Councils and self-providers. This part of the exercise is to be undertaken during the Phase 2 of the programme. 7. OVERALL CONCLUSION The current (Phase 1) NRA work has shown its value, among others by: Showing the trends in water uses. Trend analysis act as an early warning system among water users. Showing the macro-economic contribution per water unit of the different economic sectors. This allows policy makers to make a more- informed choice regarding the allocation of water to different economic activities. Comparing Botswana s water use with that of other countries. Lessons may be learnt from other countries or vice versa. The interpretation of the differences in the water use per person and the average Pula of GDP per water for the 3 countries should, however, be done cognizant of the socio-economic dimensions pertaining in each country. (Ceteris paribus assumption); Showing the extent to which losses are monitored among suppliers; Showing that given the inadequate data on costs and revenues, it is difficult to assess to what extent the pricing principles are being achieved. The analysis performed forms the initial Phase 1 in the preparation of Water accounts under the Natural Resource accounting programme. There is scope for more detailed analysis of the current data and for more analysis of the additional data that are hopefully to be obtained in future Phase 2. In addition, the current accounts are still incomplete as stock accounts are missing and monetary accounts based on costs and benefits are incomplete. Moreover, re-use of return flows (treated effluent) is not considered. The issues will be dealt with during the Phase 2 of the programme.

7 BOTSWANA S NATURAL RESOURCE ACCOUNTS: PHASE 1 REPORT OF THE WATER ACCOUNTS 1. BACKGROUND A Natural Resource Accounting Programme has been established to assess the economic value of Botswana s resources and their economic use. The Botswana Natural Resource Accounts (NRA) follow the method established by the United Nations and include water, minerals, livestock and wildlife, with other resources to be added later. Water accounts are especially significant because water is one of the critical limiting resources to development in Botswana, as it is in much of southern Africa. It is widely recognized that, with increasing resource scarcity, it is essential that economic considerations be systematically incorporated into resource allocation decisions. The environmental economic review of the National Conservation Strategy Action Plan (Arntzen and Fidzani, 1998) recommended that unless social or environmental reasons require otherwise, natural resources should be used for the most economically productive activities. This requires regular assessment of sectoral comparative advantages of resource use in terms of, for example, value-added or employment creation. The environmental economic review recommended the construction of NRA because they would provide the assessment of comparative advantage and could monitor trends in resource stocks and flows. During a pilot phase of the NRA Programme carried out by NCSA and CSO, partial accounts for water were compiled but they were not sufficiently detailed for economic analysis. Draft accounts for water have been constructed and are presented here for review. Water accounts have been constructed for each of the known major categories of water supply: Water Utilities Corporation (WUC), Department of Water Affairs (DWA), District Councils (DC), and other users who are mainly self-providers, as well as for different water sources: groundwater, dams, and rivers for the years Where possible, data are also compiled by administrative district. These accounts were discussed at a workshop of data providers and experts in order to: 1. Identify any serious errors or omissions in the water accounts (for example, wildlife and ecological water requirements are not yet included); 2. Discuss ways of improving availability and accessibility of data in required format for purposes of constructing NRAs; 3. Discuss the policy implications of the water accounts and how to improve the usefulness of the water accounts for policy-makers; 4. Discuss priorities for further work on the water accounts. The comments of the workshop have been incorporated in this report either by changes in or addition to text or as recommendations for further NRA work and research. 1.1 STRUCTURE OF THE REPORT This report is structured as follows: Section 2 gives a brief description of the framework for the water accounts.

8 Section 3 provides an overview of water use from 1993 to 1998, describing the use of water for different economic activities including losses in the distribution system The economic contribution of water use from each sector is described in section 4. Section 5 includes a comparison of water data for Namibia and South Africa. Section 6 discusses the costs of providing water and revenues received by each institution. Conclusions and suggestions for further work made by the workshop participants are presented in section 7. This is followed by Appendix A with an in-depth discussion of the data sources, the methods used to construct the accounts and the limitations of the data. Appendix B provides the more detailed water accounts. 2. FRAMEWORK FOR WATER ACCOUNTS AND SUMMARY OF DATA SOURCES A detailed discussion of the framework for the water accounts was provided in Framework for Natural Resource Accounts in Botswana, (National Conservation Strategy Agency and Central Statistics Office, Discussion Document No.1 May 1998). The most important points relevant to the water accounts are repeated here Type of accounts The accounts consist of stocks of water and use accounts for water. Accounts are constructed first in physical units (cubic meters) and, where feasible, in economic units, as a measure of the value of water. The stock accounts include information about annual quantities of water stored in dams, annual runoff to rivers, and estimated groundwater reserves, with supplemental data used as indicators such as annual rainfall. The stock accounts have not yet been developed and are not discussed here. The use accounts include the use of water for economic activities and by households according to different types of water. Different types of water are classified on the basis of natural source and institutional source. The physical accounts have been completed and are discussed here. Valuation of water is a difficult process; however, some preliminary work has been done to assess the economic benefits from water use in each sector of the economy, such as the amount of national income and employment generated by the use of water in each sector. The flow accounts as presented in this paper were discussed at the data providers workshop Classification of water supply In order to represent both environmental and socio-economic characteristics of water, the use accounts disaggregate water into a classification based on both natural source (four categories) and institutional source (four categories). There are four sources of water supply, which vary in terms of location, renewability, quality, and reliability: Groundwater, which is found throughout the country, though varying in availability and quality. For groundwater, a distinction is necessary between fossil and renewable groundwater. However, currently we are unable to do this; Dams, which capture seasonal surface water from rainfall, which is unpredictable in location, timing, and quantity, such as the Shashe and Bokaa Dams. Also, small farm dams, haffirs, and other water catchment structures are included;

9 River water, almost all supplied from internationally-shared rivers along Botswana s boundaries such as the Limpopo and the Chobe/Linyanti systems; direct use of treated effluent. Effluent is water that has already been used, but may be used again after treatment. The water accounts include groundwater, dams, and river water, but do not yet include the use of effluent. Effluent has been identified as a water source of great potential. For example, in 1990, Gaborone s effluent was estimated to equal nearly two-third of the capital s water consumption (SMEC et al., 1991). It is expected that the forthcoming Sanitation and wastewater Policy and National Sanitation and Waste Water Master Plan will analyze the amount of effluent available and best ways of re-using and recycling this source of water. These efforts will provide data that can be incorporated in future NRA work on water. In addition, a number of different institutions provide water. Some of these institutions rely on relatively large-scale, technologically sophisticated infrastructure for collection and longdistance water distribution networks, while others rely mostly on local, small-scale infrastructure such as local boreholes and small dams. The former generally institute formal systems of water metering and serve urban areas and the important mining, industrial and commercial activities whereas the latter generally serve the needs of the rural population and may not keep formal records about water use. The Botswana water supply institutions include: Water Utilities Corporation (WUC), which provides water to six urban areas; Department of Water Affairs (DWA), which provides water to seventeen major villages; District Councils (DC), which provide water to more than 200 small villages; Users, who mainly provide their own water outside the other three institutions. Such self-providers are mainly found in the mining, livestock, irrigation, and wildlife subsectors. All the providers of water are included in the accounts. Transfers of water from one institution to another, for example, sales from Jwaneng mine to WUC for distribution to Jwaneng, are not traced in the water accounts. Only the institution that provides the water directly to the end-user is identified at this time. Ideally, a distinction needs to be made between international water sources (or shared resources) and domestic ones. The former is shared with other countries, and their use is subject to international treaties and negotiations. In contrast, use of the latter is an entirely domestic affair. At this stage, this distinction could be only partly made. Perennial rivers are all international and dams on ephemeral rivers are all domestic. Most groundwater is also domestic, but there are aquifers that cross national boundaries; these have not yet been identified, but will be considered in future work. In summary, the Water Accounts presented here are disaggregated by three types of natural sources and four types of institutional sources. In addition, the spatial aspect of water is important because water availability and quality varies a great deal by region. To the extent possible, use of water is also identified by administrative district so that district-level water accounts can be constructed.

10 2.3. Classification of water users The purpose of environmental accounts, and water account in particular, is to link information about water use to economic information contained in the National Accounts. This link is established by using the same classification of economic activities for both water accounts and National Accounts. The National Accounts has a well-established classification of economic activities, which has been used to classify water use (this classification of economic activities is shown in Table B1 in Appendix B). Some important uses of water are not reflected in the national accounts' classification because these uses do not have a corresponding economic activity. Notable among these uses are various ecological water requirements. The ecological water requirements include direct use of water by wildlife, as well as in-stream flow requirements of rivers, and the water requirements needed to maintain a healthy ecosystem in areas like the Okavango delta. These uses of water have not been included at this time because of a lack of data, but efforts will be made to include them in future work Data sources and major data problems A detailed discussion of data sources for the water accounts, and the problems with the data is provided in Appendix A. A summary is provided here in Text Box 1, in order to guide readers in interpretation of the analysis that follows, and to focus readers' attention on the major limitations. Two indicators are used to rank the relative importance of each major provider i.e. the percent of total water provided, and the percent of total groundwater used from these sources. It may be more important, for example, to address data problems with sources that provide a relatively large share of water, such as District Councils, than DWA. Similarly, because of the concerns over use of non-renewable groundwater, it may also be more important to focus on improving the data for sources that relies heavily on groundwater, such as District Councils and self-providers (mining, irrigation, livestock ranches etc). WUC provides a significant amount of water, 19% in 1998, but does not rely on groundwater. DWA provides a relatively small amount of total water use, 9% in 1998, and accounts for 13% of groundwater used. DC provides 17% of water use, but account for a large share of groundwater, 31%. Finally, self-providers account for the largest share of water use and also the largest share of groundwater, 55% and 56%, respectively. 3. OVERVIEW OF WATER USE IN BOTSWANA (1993 AND 1998) This section describes the use of water over the period, 1993, which is the first year for which complete water use accounts were constructed, to 1998, the most recent year for the accounts. In most instances in this report, only figures for 1993 and 1998 are presented in order to make the presentation more clear without cluttering it with unnecessary detail. More detailed accounts are provided in Appendix B. This section considers: national trends in water use; use by different economic sectors of the economy; Un-accounted for water and losses in delivery of water to the end user.

11 Text Box 1 Summary of data sources and data problems A. WATER UTILITIES CORPORATION RELATIVE IMPORTANCE Percentage Use in 1998 of Total water: 19% of Ground water: 0% Data Source: Unpublished data base of billing records; Unpublished data base of water use and tariffs by customer type and by tariff band Major Problems: Discrepancies between billing records and published figures for water use B. DEPARTMENT OF WATER AFFAIRS RELATIVE IMPORTANCE Percentage Use in 1998 of Total water: 9% of Ground water: 13% Data Source: For each major village, unpublished databases with the following information: Production, Consumption, and Losses; Categorized Consumption; Expenditures and Revenues; Major Problems: Missing data for water use in many villages; Extensive missing data for Expenditures and Revenues Incomplete record of Overhead & Maintenance costs in the Expenditure database Revenues; No information about capital costs. C. DISTRICT COUNCIL WATER FOR SMALL VILLAGES RELATIVE IMPORTANCE Percentage Use in 1998 of Total water: 17% of Ground water: 31% Data Source: Estimated on the basis of per capita daily water use in several hundred small villages. Per capita water use derived from one month of metered water use in each village, used to derive average daily use per person. Major Problems: No information about how much water is used for domestic consumption, schools & clinics, offices, local government, livestock water, and other uses. No information about costs or revenues. D. SELF-PROVIDERS (mainly mining, irrigation, livestock watering industries) RELATIVE IMPORTANCE Percentage Use in 1998 of Total water: 55% of Ground water: 56% Data Source: Livestock: estimated on the basis of numbers of livestock and daily water requirements. Irrigation: using same figure as used in Water Master Plan for Diamond mining: reported use from Debswana from only. Other mining: copper/nickel & soda ash from WUC records; coal from Water Master Plan. Major Problems: No reliable livestock figures from Actual livestock water use unknown. Mix of groundwater and other sources of water for livestock unknown. Actual irrigation water use unknown. Diamond mining water use before 1996 unknown. No information about costs of providing water for any users.

12 3.1. National Trends in Water Use Total water use in Botswana has increased only by 5% from about 143 million cubic meters in 1993 to 150 million cubic meters in 1998, (Table 1). This increase is relatively low considering the much higher rates of economic growth and population growth over the period (Figure 1). Per capita water use actually declined and the macro-economic efficiency of water use increased. Substantial increases in GDP were achieved without requiring much additional water. This issue is further discussed in section 4. The trends in water use are summarized in Tables 1-2 and Figures 1-3. Part of the explanation for the slow growth of total water use is seen by examining the water accounts for each of the institutions supplying water. Water use from WUC was fairly constant from , but then shot up a great deal in 1997 and Water use from DWA increased a great deal from 1993 to 1994, then remained relatively constant for three years until 1997 and 1998, when it began to rise again. Water supplied by District Councils to the small villages has increased at a steady rate, but this is the result of estimation methods, which assumed that water use grows in the same proportion as the rural population (see Appendix A for further discussion). Water used by self-providers also fluctuated a great deal, falling from 87 million cubic meters in 1993 to 82 million by This decline is due to a fall in livestock numbers as well as improvements in water efficiency in diamond mining. The formal water supplying institutions WUC, DWA and DC supplied 45% of water used in 1998, up from 39% in 1993 (Table 1). The quantity of water attributable to self-providers either declined or remained constant, reducing their shares of total water (Figure 2). Botswana relies on groundwater for roughly 60% of its water needs, but dams are slowly contributing a larger share (Table 1, Figure 3). While the amount of groundwater and river water remained virtually unchanged between 1993 and 1998, the amount supplied from dams increased, changing its share from 19% in 1993 to 23% in Most of the ground water is used in rural areas, either in small villages for domestic consumption, or by self-providers, mainly for diamond mining and livestock watering. Most of the river water is used for irrigation, though it is an important source of water for some villages as well.

13 Table 1. Water use by supplying institution and natural source, 1993 to 1998 A. Millions of cubic meters 1993/ / / / / /99 Use by natural source Groundwater Dams River Total, all sources Use by institution and natural source WUC Dams DWA Groundwater Dams River Total DWA DC Groundwater Selfproviders Groundwater Dams River Total B. Percentage distribution by institutional source 1993/ / / / / /99 WUC 16% 16% 16% 16% 17% 19% DWA 6% 7% 7% 8% 8% 9% DC 17% 17% 17% 18% 17% 17% Self-providers 61% 60% 60% 58% 58% 55% Total 100% 100% 100% 100% 100% 100% C. Percentage distribution by natural source 1993/ / / / / /99 Groundwater 60% 60% 60% 59% 59% 57% Dams 19% 19% 19% 20% 21% 23% River 21% 21% 21% 21% 20% 20% Total 100% 100% 100% 100% 100% 100% Source: See text for discussion of data sources.

14 Figure 1. Index of volume of water used and intensity of water use by population and GDP, 1993 to 1998 (1993 = 1.00) Volume of water Per capita water use GDP per m3 water / / / / / /99 Source: Volume of water, Table 1. Population and GDP Data from CSO. Figure 2. Water supplied by major institution in Botswana, 1993 to / /99 millions of cubic meters W U C DW A D C Self-providers Source: Table 1

15 Figure 3. Water use by natural source in Botswana, 1993 to milliopns of cubic meters of water / /99 - Source: Table 1 Groundw ater Dam s River From trends at the national level, one would conclude that the continued heavy reliance on groundwater warrants close monitoring of the balance between recharge and use. This monitoring is especially important in the rural areas, where use of groundwater is high and alternatives to local groundwater supplies are limited. The trends also indicate an increasing role for formal sector water supply WUC, DWA, and DC. This can be a positive step because it provides the institutional setting for better water monitoring, control and management, especially of groundwater. However, self-providers still account for most water use, and while some of this is monitored (the mining industries file annual reports on their water use), much of it is not, especially with irrigation and livestock water use. The decline in water use can easily lead to complacency among water managers. While it may be a reason to be optimistic, such complacency is not justified until a detailed analysis of water use over a much longer period (at least 10 years) has been made. This is an area for further research Water Use by Economic Sectors Agriculture is the major user of water in most countries and Botswana is no exception. 50% of water use in 1993 was for agriculture, about 35% in the traditional sector, dominated by livestock, and the other 15% in commercial agriculture which is mostly irrigated crop farming (Table 2). However, it should be noted that while high, Botswana's use of water for agriculture is considerably lower than in its neighbors Namibia and South Africa (see Section 4). The difference is primarily because Botswana has not developed extensive irrigated agriculture. This may change in future if plans to expand irrigation are implemented. Agriculture's share of total water use declined to about 45% in 1998 due to a decline in cattle numbers. Agricultural water use will, of course, rise in absolute amounts if livestock numbers increase again; its share of water is likely to decline as the use of water by other sectors increases. Livestock watering (both traditional and commercial) in rural areas is still the single largest user of water, although its share declined from 37% in 1993 to 32% in Though Botswana does not undertake much irrigation compared to other countries in the region, irrigation farmers account for a significant share of water (9%).

16 Table 2. Water use by detailed economic sector in 1993 and 1998 cubic meters Percentage 1993/ / / /99 All agriculture 71,693,610 67,336, % 44.9% Traditional agriculture 49,879,471 45,728, Commercial agriculture 21,814,139 21,608, All mining 16,207,370 15,546, Diamond 12,038,060 9,772, Copper/nickel 3,825,441 5,362, Coal+ other 343, , Manufacturing 2,215,297 3,023, Food and beverages 1,488,529 1,543, Other manufacturing 726,768 1,479, Water & electricity 1,539,260 1,536, Construction 309, , Services 3,124,052 3,244, Trade, hotels & restaurants 1,253,589 1,247, Transport, Communications 166, , Insurance, Banking, Business 472, , Social and personal services 1,231,766 1,249, Government 7,221,229 9,825, Central 5,920,379 8,170, Local 1,300,851 1,654, Households 40,526,175 49,144, Urban 10,817,245 13,574, Peri-urban 5,446,493 9,392, Rural 24,262,437 26,178, Total 142,836, ,844, Source: See text for discussion of sources. Mining, the mainstay of the Botswana economy, is a significant user of water, accounting for 10% in 1998, but this was lower both in relative and absolute terms compared to In 1993, total water use for mining was 16.2 million cubic meters (11.4% of total use), but this had declined to 15.5 million cubic meters by The reduction in water use was due entirely to water conservation measures introduced in diamond mining, which reduced water use by 19% between 1993 and During the same period, water use for all other mining increased; the use of water for copper/nickel mining increased by almost 40%, even though physical production did not increase much. Water use by service industries hardly changed between 1993 and 1998, but the use of water by government increased by 30%, increasing its share of total water use in the economy from 5% to 6.6%. Following agriculture, domestic use by households constitutes the second largest user of water, and one that is rising faster than most sectors of the economy. Though

17 total water use increased by only 5%, household water use increased by 21%, increasing its share from 28% to 33% in The shares of domestic water use and of population among the three different categories of settlements (towns, major villages and small villages) are fairly similar (Figure 4). This result is surprising because in most countries urban water use is much higher than rural water use. However, the figure for small village water use is probably too high since, as mentioned at the beginning of this report, all water use in small villages is attributed to households, and at least part of this water use is for other economic activities. The figures for towns and major villages are more comparable, and here there is a difference in per capita annual water use registering 36 cubic meters per person in towns and only 26 cubic meters in major villages. Agriculture, households, and mining continue to be the largest sources of water demand. The rapid increase in household water use from 1993 to 1998 is potentially a problem if demand continues to increase at such a rate. However, the dramatic reductions in water use by diamond mining demonstrate that there is considerable room for water conservation measures. Steps should be taken to expand water conservation measures to other major users, especially households. Figure 4. Percentage shares of population and of domestic water use by size of settlement in % 50% % of population % of domestic water use 40% 30% 20% 10% 0% Towns Major Villages Small villages Source: water use from Table 2, population figures from CSO population projections Unaccounted for Water and Supply Losses All providers suffer significant losses of water between the starting point of the supply system and the end user, which are called unaccounted for water. Losses occur because of leakages within the system, illegal use of water from the system, or poor monitoring and meter reading. Losses due to system leakage are not directly attributable to the use of water by any specific economic sector. Unaccounted for water due to illegal use or poor

18 monitoring may be for an economic purpose, but there is not sufficient information to assign it to any particular sector. Although unaccounted for water cannot be attributed to specific economic activities, it is important to monitor this figure because it represents a drain on scarce resources as well as a potential source of additional supply often at a relatively low cost. Few institutions regularly measure and report unaccounted for water. WUC started to estimate this figure only in 1998; a figure for Gaborone is also available for 1997 (Table 3). After a special study of unaccounted for water in Gaborone in 1997 then estimated at about 30% WUC undertook a programme to reduce physical leakages in the city system. The result was a dramatic reduction, from 30% to 14% in 1998 and 9% in 1999 (pers. com. WUC). DWA estimates losses annually as the difference between production of potable water and what is metered at a customer's premises. Though this figure is officially called loss, it includes both leakages and unmonitored water use, and hence conforms to the definition of unaccounted for water. To be consistent with DWA terminology, it is referred to as losses in this report. The average rate across the 17 major villages served by DWA has declined from 28% in 1993 to 24% in However, there is a tremendous range of loss rates for the different villages supplied by DWA. In 1993 the lowest reported loss rate was 6% for Letlhakane and the highest was 49% in Mogoditshane. In 1998 the spread was smaller, 12% in Letlhakane and 35% in Ramotswa. District Councils and self-providers do not provide estimates of losses. Table 3 further shows that only 28% of the total water use are checked for losses, i.e. WUC and DWA only. Table 3. Reported losses of water (in percent) WUC Gaborone Na Lobatse Na Na Jwaneng Na Na F/town Na Na S/Phikwe NA NA 10 9 Sowa Na Na DWA Average Na Village with highest loss Na Village with lowest loss rate Na DC Na Na Na Na Self-providers Na Na Na Na Note: Na = not available Source: WUC; DWA, Annual Reports discussed in this text.

19 In conclusion, the loss rates are relatively high, except in Gaborone and Selebi-Phikwe. Although information about losses is only available for WUC and DWA, there is no reason to assume that loss rates are much lower for the other providers. If we assume that losses of other providers are similar, the country s total loss rate is 15.7%. Clearly, there is need for more comprehensive monitoring of losses by all providers, for example as part of environmental auditing. 4. ECONOMIC BENEFITS OF WATER USE BY SECTOR In a water-scarce country, the provision of water must always be considered in terms of the socio-economic benefits a particular user contributes to national development. It is not easy to measure a precise economic value for water, but there are two commonly used indicators of the socio-economic benefits from water use: the national income and the employment from the use of water in each sector. National income is measured as value-added by a sector's contribution to Gross Domestic Product (GDP). In many countries now, "environmental economic profiles are constructed that compare the percentage use of water by a sector to its percentage contribution to national income and employment. Environmental economic profiles can be useful in comparing the allocation of water across sectors, and in "benchmarking" industry s performance. Benchmarking tells an industry whether it is improving its water efficiency over time, how it compares to other industries, how it compares to the same industry in other countries, and can be used by individual companies to see how they compare to the average performance for their industry. Figure 5 below presents the environmental economic profile for water in Botswana in Of the major sectors, Agriculture, using 50% of all water in 1993, contributed only 4% of national income (value-added) and 2% of formal sector employment. The socio-economic contribution of Agriculture in terms of income and employment is the lowest of all sectors relative to the share of water it uses. This sort of imbalance between water demand and socio-economic benefit is seen in many countries, due to the nature of agriculture as a water-intensive activity. However, the employment figures for Botswana understate the social importance of agriculture because only formal-sector employment is considered. Many more people rely on agriculture for at least part of their livelihoods. When employment is adjusted to include full-time farmers (Figure 6), the share of employment provided by agriculture increases to over 30%, still less than the share of water used, but clearly of enormous social significance. Mining makes a strong economic contribution in terms of income (35%), but not employment (under 4%), relative to its water use (10%). All of the remaining sectors account for higher shares of income and employment compared to their shares of water used. The socio-economic contribution is especially high for services and services like banking and insurance will use even less water than other services such as hotels and restaurants.

20 Figure 5. Percent of water use, national income, and formal employment contributed by major economic sectors in % 50% 40% Water use National income Employment 30% 20% 10% 0% Agriculture Mining Manufact. Water & electricity Construction Services Government Source: Water data from sources described in the text; employment and value-added data from the Central Statistics Office, Statistical Abstract Figure 6. Percent of water use, formal employment, and employment including fulltime farmers in % 50% 40% Water use Formal employment E m ploym ent inc. full-tim e farm ers 30% 20% 10% 0% Agriculture Mining Manufact., U tilitie s, Construction Services Government

21 Note: Figures for full-time farmers only available for The same figure was assumed for Source: Water data from sources described in the text; employment from the Central Statistics Office, Statistical Abstract 1998, full-time farmers from the 1991 Population Census. The economic sectors shown in Figure 5 are highly aggregated and, consequently, can hide a great deal of variation within a larger sector. Table 4 shows more detailed information about economic contribution by economic sectors for 1993 and Figure 7 shows this detail for the Mining sector. Within Agriculture, the traditional sector makes a greater economic contribution, relative to the share of water it uses, than the commercial sector. This is due to the fact that livestock dominates the traditional sector while the commercial sector is dominated by irrigation, which uses much more water in production. Within the Mining sector, diamond mining, not surprisingly, makes a stronger economic contribution compared to the other sectors. The Manufacturing sector shows considerable variation the waterintensive food and beverage sectors generate less income and employment relative to the water they require compared to other manufacturing sectors. The Service sectors also show variation in water use intensity and economic contribution. Figure 7. Water use and contribution to national income by mining in % 35.0% 30.0% 25.0% % of Value-added % of water use 20.0% 15.0% 10.0% 5.0% 0.0% All mining Diamond Copper/nickel Coal+ other Source: Water data from sources described in the text; value-added data from the Central Statistics Office, Social Accounting Matrix for 1993/94. Table 4 also provides a comparison of the years 1993 and Only the major sectoral groupings can be compared because information about value-added and employment by all other economic sectors is not yet available. Manufacturing, Government, and Households have all increased their shares of water used while Agriculture and Mining have decreased their shares. The declining use of water in Agriculture reflects a parallel decline in the income and employment in that sector, whereas the decline in water use for Mining has been accompanied by little change in income and employment. The declining use of water in

22 Mining has been achieved, in part, by greater efficiency in the use of water in diamond mining. Figure 8 summarizes the comparison of the contribution to national income relative to water use by economic sector, and how it has changed over time. The level of value-added was divided by the total water use in each sector. For this comparison, the value-added for both years is given in the same year's prices, 1993/94 prices, in order to adjust for inflation, which would otherwise bias the comparison in favor of This calculation provides a figure for the Pula of national income produced per cubic meter of water input by each sector in 1993 and Table 5 reports these values, along with the percentage of water used by each sector in Table 4. Percent distribution of water use, national income, and employment by economic sector in 1993 and 1998 Economic Activity 1993/ /99 Water National Employment Water National Employ- use income use income ment Agriculture 50.2% 4.4% 2.1% 44.9% 2.8% 1.7% Traditional agriculture 34.9% 3.8% Commercial agriculture 15.3% 0.4% Mining 11.3% 35.7% 3.7% 10.4% 35.0% 3.6% Diamond 8.4% 32.5% Copper/nickel 2.7% 1.9% Coal+ other 0.2% 1.2% Manufacturing 1.6% 4.6% 9.1% 2.0% 4.7% 10.0% Food and beverages 1.0% 1.8% Other manufacturing 0.5% 2.9% Water & electricity 1.1% 2.2% 1.1% 1.0% 1.9% 1.1% Construction 0.2% 6.5% 12.6% 0.1% 5.9% 9.4% Services 2.2% 26.4% 36.2% 2.2% 30.3% 32.4% Trade, hotels & restaurants 0.9% 8.3% 18.4% 0.8% 11.8% 18.0% Transport, Communications 0.1% 3.7% 4.0% 0.1% 4.1% 3.7% Insurance, Banking, Business 0.3% 10.1% 8.0% 0.4% 10.5% 7.1% Social and personal services 0.9% 4.3% 5.8% 0.8% 3.9% 3.6% Government 5.1% 15.6% 35.2% 6.6% 16.0% 41.8% Central 4.1% 13.3% Local 0.9% 2.3% Households 28.4% 32.8% Source: See text for discussion of sources.

23 Figure 8. Water use and contribution to national income by economic sector in 1993 and 1998 pula per meter of water in constant 1993/94 prices Pula VA/m constant price pula VA/m3 Agriculture Mining Manufacturing Water & electricity Services Government GDP Note: Contribution to national income is measured as the value-added contributed to GDP by each sector. VA = value added. Source: Table 5. Overall, the economy is producing higher levels of income for a given amount of water used. The value rose from nearly P 77 to P 96, an increase of 26%. This increase in water productivity may have been achieved in two ways: through the introduction of water conservation measures that reduce the amount of water needed for a given level of production; or through a change in the structure of the economy, substituting activities that are relatively high-value per cubic meter of water input for those that are relatively low-value. Until more detailed economic information is available for 1998, which would allow further dis-aggregation of economic sectors, it is difficult to determine the extent to which each factor has contributed to this overall increase in water efficiency. Nevertheless, some observations can be made as follows: Economic efficiency of water use has two components that work at different levels of the economy. At the level of the company or industry, companies can take measures to conserve water, producing the same quantity of output, but with less water. This is water efficiency at the industry-level. At the macro-economic level, water efficiency is achieved by directing

24 Table 5. Water use and contribution to national income by economic sector in 1993 and 1998 Pula of value-added per m3 of water used in constant 1993/94 prices Percentage of water use in Agriculture % Mining % Manufacturing % Water & electricity % Construction 2, , % Services , % Trade, hotels & restaurants , % Transport, Communications 2, , % Insurance, Banking, Business 2, , % Social and personal services % Government % Households na na 32.8% Average for GDP % Note: Households do not generate value-added so this calculation cannot be carried out for this user. Source: Water data from sources described in the text; value-added data from the Central Statistics Office, Statistical Abstract Production toward those services that generate higher levels of income and employment per cubic meter of water used for example, services instead of agriculture. Over time, water efficiency can be improved in both ways: companies in each industry need to conserve water (industry-level water efficiency) and the structure of the economy should change to favor less water intensive industries (macroeconomic water efficiency). Two sectors are particularly important in determining water efficiency and the value of GDP per cubic meter of water input: Agriculture and Mining. Agriculture is important because it is the major user of water, roughly half, although its contribution to GDP is small. Mining is important because it is the largest source of GDP, as well as a major user of water. In Agriculture, the economic contribution of water per cubic meter of water declined between 1993 and 1998, from P 6.75 to P This change resulted mainly from the decline in livestock, a more valuable use of water than irrigation, as was seen in Table 2. Such a decline would actually work towards reducing the overall water efficiency in Agriculture and in the overall economy (lowering macroeconomic water efficiency) by reducing the GDP per cubic meter of water. However, there are several counteracting trends in the economy. The economic efficiency of water use in Mining increased by an impressive 25%, from P 244 per cubic meter of water in 1993 to P 302 in This is due in large part to the water conservation efforts by Debswana. A recent study of water demand management identified Debswana as one of the leaders in water conservation (Arntzen et al., 1999). Table 2 and detailed tables in the Appendix show that the use of water by diamond mining has declined, while for the other sectors it has increased