The value of Namibia's forest resources: Preliminary economic asset and flow accounts

Abstract

A national forest inventory completed in 2004 enabled the development of preliminary forest resource accounts for Namibia. The total woody resources volume was estimated at 257 million m³. Forest products used for fuel, poles, timber and non-timber were included in the accounts, but charcoal production on private land was excluded. Forest use directly contributed some 3 per cent of the gross national product, N$1 billion (US$160 million), and the total direct and indirect economic impact of forest use in the broader economy was some N$1.8 billion (US$288 million). The standing forest assets had an estimated capital value of N$19 billion (US$304 million), comparable with the values for fish, minerals and wildlife. On a national level, stocks are underused. However, some localised over-harvesting occurs, which might be ameliorated through community forest management and trade in products, but capture of resource rent by government should be restricted to commercial forest use.

Keywords:

• forests
• use value
• asset value
• resource accounts
• Namibia

1. Introduction

This paper describes the preliminary development of natural resource accounts for forest resources in Namibia. The work follows and completes that of Björkmann (1999), who developed a partial set of accounts. Forest accounts estimate the economic value of forest use and standing forest assets and provide guidelines for forest sector policy.

True forest is absent from Namibia, and forest resources are here defined as all woody plants that occur in the woodlands and shrublands (savannas) of the country. Namibia embraces some 824 000 km² on the southwestern coast of Africa, and has a human population of 1.8 million. The climate ranges from extremely arid in the west and south, to semi-arid in the northeast. Woody resources are extremely sparse in the arid desert environment, but increase in density toward the northeast, through shrubland and savanna habitats commonly dominated by Acacia, Terminalia and Colophospermum mopane, to woodland where Burkea africana, Combretum and Baikiaea plurijuga tend to dominate. There are almost no planted forests, and the accounts deal exclusively with natural forest resources. Most of the arid and semi-arid shrubland and savanna in the centre and south of the country is private land under commercial livestock production, while most of the semi-arid savanna and woodland in the north and northeast of the country is communal land under traditional pastoralism and agro-pastoral land use.

Natural resources have several types of economic value, and in resource economics these values are commonly classified under ‘total economic value’. This embraces direct use values, indirect use values and non-use values. Direct use values derive from the direct use of the resource, i.e. in production of tangible goods, usually with market value; indirect use values from the resource's value in ensuring ecological function, such as watershed conservation; and non-use values from the value of conservation of the resource either for future use (option value), for its mere existence (existence value), or to bequeath to future generations (bequest value). The preliminary forest accounts deal exclusively with direct use values.

The direct use values derived from the use of Namibia's natural shrubland and woodland resources come from harvesting fuel wood and poles for constructing houses and fences. The wood harvested is mostly consumed directly by rural households, but is also for limited sale in urban areas. In the past, commercial harvesting of saw timber took place in the woodlands of the northeast. Here, resources of saw timber, including species such as B. plurijuga, Pterocarpus angolensis, Guibourtia coleosperma and B. africana, exist on Kalahari sand. No commercial extraction of these species currently takes place. Forest use value also comes from other plant products, most of which are harvested for home consumption by rural households. These non-timber forest products (NTFPs) include plant products for craft production (carving, basket-making); plant products for food, medicine and cosmetics; and grass for thatching. A new and increasingly important use of forest resources is the production of charcoal from encroaching shrubs on private land. Due to lack of available data, the charcoal industry has been excluded from the preliminary forest accounts. The forest accounts also do not include the use of woodlands and savannas for livestock grazing or for their wild fauna. These activities are considered best treated in separate accounts.

The forest accounting activity forms one component of a broader national natural resource accounting programme being undertaken by the Environmental Economics Unit of the Ministry of Environment and Tourism. This programme extends the conventional macroeconomic national accounts through the development of satellite asset accounts for natural resources such as fish, forests, minerals, water and wildlife. While the use of these natural resources has been included in the conventional national accounts, the resources have not been accounted for as assets. National accounts have historically only incorporated man-made or owned assets. However, exclusion of natural assets from the national accounts prevents sound planning for sustainable development.

In natural resource accounting, the natural assets are valued in two ways. First, the annual contribution of the resource to the national income, in terms of direct use values, is measured in a production, or flow, account. Second, the value of all existing stocks of the resource is estimated in an asset account. Here the value of the stock, as a national capital asset, is measured in terms of its potential to generate resource rent (also known as ‘economic rent’ or ‘excess profit’) in the future.

2. Methods

Table 1 summarises some of the criteria and assumptions used in developing the forest accounts. The approach and methods used to derive these criteria and assumptions, as well as the accounts themselves, are described below. Monetary values in this paper are given as Namibia dollars (N$) at 2004 prices. In 2004, N$1.00 was equal to R1.00 (South African rand), P0.75 (Botswana pula) and US$0.16 (US dollar). The table shows clear differences among the Regions for usage rates of woody biomass. This is for a range of reasons discussed further in this paper. Data on Namibia's forests remain poor, necessitating the use of assumptions, and disaggregated data on NTFPs, a crucial element of rural livelihoods, are largely absent.

Table 1: Some key criteria and assumptions applied in the forest accounts for Namibia, 2004

Measure	Units	Forest products
		Fuel wood	Poles	Saw timber	NTFPs^a
% stock physically usable	Per cent	90	15	1	–
Yield/physically usable stock	Per cent	7	7	6	–
% annual yield currently used	Per cent	6	12	0	–
% annual yield economically usable	Per cent	16	27	50	–
Yield (mean annual increment) by Region
Caprivi	m^3/ha/year	1.7310	0.2885	0.0821	–
Erongo	m^3/ha/year	0.0036	0.0006	–	–
Hardap	m^3/ha/year	0.0036	0.0006	–	–
Karas	m^3/ha/year	0.0014	0.0002	–	–
Kavango	m^3/ha/year	1.2960	0.2160	0.0199	–
Khomas	m^3/ha/year	0.0090	0.0015	–	–
Kunene	m^3/ha/year	0.0054	0.0009	–	–
Ohangwena	m^3/ha/year	1.4400	0.2400	0.0200	–
Omaheke	m^3/ha/year	0.0900	0.0150	–	–
Omusati	m^3/ha/year	0.1447	0.0241	–	–
Oshana	m^3/ha/year	0.0405	0.0068	–	–
Oshikoto	m^3/ha/year	0.6178	0.1030	–	–
Otjozondjupa	m^3/ha/year	0.2106	0.0351	–	–
Consumption per household	m^3/year	4.60	4.37	–	–
Price per m^3	N$(2004)	634	529	5753
NTFP value/fuel wood value	Per cent	–	–	–	64
Direct value added	Per cent/output	94	89	66	68
Total value added	Per cent/output	149%	149	114	149
Income multiplier	Factor	1.71	1.71	1.59	1.71
Resource rent	Per cent/output	84	73	41	68
^aNTFPs used for food, medicine, cosmetics, craft-making and thatching.

2.1 Approach

The asset and flow accounts were developed in accordance with the standardised methodology for natural resource accounting – the Integrated Environmental and Economic Accounting/Integrated Environmental and Economic Accounting (IEEA) Manual – developed by the United Nations (UN, 2000) and later refined by UN et al. (2003) and described in manuals by Eurostat (2002) and Lange (2004a). The IEEA Manual was developed to complement the conventional, internationally adopted System of National Accounts, used to measure economic performance in most countries around the world (CEC et al., 1993). Conventional national accounting incorporates capital accounts, but tends to restrict these to assets that are owned or man-made. The IEEA Manual, on the other hand, aims to include accounts for resources that are not man-made, such as natural forests, fish and wildlife, in the national economic data records and planning.

Our approach involved first developing physical accounts for national natural forest assets. These detail the standing volumes of woody resources in each administrative Region of the country. We then detailed the current annual use of these resources in flow accounts, which present the volumes and economic characteristics of this use. Flow accounts measure use in terms of output, contribution to gross national product (GNP) and employment, in conformity with the System of National Accounts. The physical accounts are then valued in order to produce monetary asset accounts, so that changes in the capital value of forests can be measured. As stated above, the value of natural assets is measured as the resource rent that can be generated from their future use.

The valuation of renewable natural resource assets such as forests, fish and wildlife can be done in several ways (UN, 2000; UN et al., 2003). The most appropriate, the Net Present Value Method, estimates the present value of all future returns in resource rent from the use of the resources. In the absence of realistic data on future use and discount rates, other – less satisfactory – alternatives are often employed, such as the Net Price Method or, as it is called in the case of forests, the Stumpage Value Method.

This study used the Net Present Value Method, with streams of rents from expected growth in use over the next 30 years. Forest stocks tend to be stable and their future use pattern predictable. With resources such as fish and minerals, future stock and use values are much less predictable and researchers who have valued these (e.g. Lange & Hassan, 2003; Lange, 2004b) have had to assume no growth in use or constant rent into the long-term future.

One important consideration in valuing natural asset stocks is that resources that are not exploitable, either for legal or economic reasons, have a zero value. We therefore valued only that portion that could realistically be brought into viable production in the future.

The accounts developed are for the year 2004. Asset accounts should include consideration of depletion, degradation, conversion and accumulation of stocks during the accounting year. The changes in volume and value of stocks can then be accounted for over time. The current use of forest products is significantly below the sustainable and economically possible potential for this use, and data are lacking to measure changes in overall stock volumes. In any case, at this stage, such changes are expected to be extremely small, relative to the overall stock.

2.2 Field survey

A specific household and focus group survey was conducted to obtain data on rural residents' use of forest resources (fuel wood, poles and NTFPs). This provided measures associated with resource use and sales, which, combined with parameters extracted from the literature, enabled computation of the accounts.

A stratified sample of 182 households in the Ohangwena, Omusati, Oshana and Oshikoto Regions of north-central Namibia was targeted. The sample was designed to cover randomised residents from each of the eight biomes present in the Regions that support human populations. The questionnaire was aimed at obtaining quantitative information on a basic household profile, including volumes of forest resources harvested, consumed and sold; prices; harvesting costs; and the relative importance of income from forest resources to household livelihoods. The questionnaire was finalised after two pilot surveys, each of which involved 10 respondents. Two graduate economists and four undergraduate enumerators, who were trained prior to the survey, were deployed. Household heads were interviewed in Oshiwambo for approximately 1 hour.

Additionally, a sample of 25 forest product traders and trader groups in the north-central Regions were targeted by means of a questionnaire to solicit information on the economics of trading in fuel wood, poles and NFTPs. This complemented the household survey and provided additional information on the size and characteristics of product market chains. The sampled entrepreneurs were from both rural and urban locations, and were both full-time and part-time traders. They operated at open markets in the urban centres of Ondangwa and Oshakati and on public roads across the survey area.

2.3 Standing woody resource volumes

The forest asset account was based on information about standing woody biomass volumes, which had been assembled by the national forest inventory subcomponent of the Namibia–Finland Forestry Programme in the Directorate of Forestry. These data are presented in various Directorate reports. The standing volumes make up the physical asset account.

Standing volumes were estimated for each Region in the country using a combination of Region-wide inventory estimates (e.g. for Omusati (Selänniemi et al., 2000a), Oshana (Selänniemi et al., 2000b), Oshikoto (Angombe & Laamanen, 2002) and Otjozondjupa (Korhonen et al., 1997)), local inventory estimates (in order of importance, Chakanga et al., 1998b, 1999; Chakanga & Selänniemi, 1999; Angombe et al., 2000, 2001; Laamanen & Angombe, 2001; Laamanen et al., 2002; Mulofwa et al., 2002; Kanime, 2002, 2004; Kanime & Laamanen, 2002, 2003; Kanime & Kakondo, 2003; Kamwi, 2003a, 2003b; Boois, 2004; Mwilima & Boois, 2004) and ecological information (estimates made in similar habitats in Botswana by Nickerson, 1984; ERL Energy Resources Ltd, 1985; Norwegian Forestry Society, 1992) to corroborate the interpolations and extrapolations. Standing volumes, specifically as regards the more valuable saw timber species that occur in the northeast, were also obtained from the inventories. Here estimates for the whole Caprivi Region were provided by Chakanga et al. (1998a), and local measures were used to interpolate and extrapolate estimates, from Chakanga et al. (1999), Angombe et al. (2000, 2001), Kanime & Laamanen (2002), Mulofwa et al. (2002), Kamwi (2003a, 2003b), Kanime & Kakondo (2003) and Kanime (2004).

2.4 Current use of woody resources

The flow account requires a measure of the current use of natural woody resources (2004). Current use has tended to be restricted to areas relatively close to human settlement. In Namibia, fairly extensive tracts of natural woodland and savanna are uninhabited, and are not currently used. For fuel wood and poles, aggregate output values were obtained by multiplying the estimates of product sales/consumption per household by the numbers of households that make use of these resources in each Region. Results from the National Population and Housing Census of 2001 (unpublished data, Central Bureau of Statistics, National Planning Commission, 2004) contain statistics on the proportions of households in each Region that use wood for fuel and poles for construction and fencing.

Measurements of fuel wood consumption are highly variable, so the annual consumption of fuel wood, per household and per capita, was calculated by averaging results from 12 studies from South Africa and six from Namibia. Besides a personal communication from S el Obeid (2005), these were extracted from Bembridge & Tarlton (1990), Ollikainen (1991) and Shackleton et al. (2004), as well as from our own survey, as described above. After two outliers were excluded, the average came to 4.60 m³ per household, or 0.88 m³ per person per annum. For pole consumption, our survey estimate was combined with data from the forest inventory (Selänniemi et al., 2000a, 2000b) to derive an average (4.37 m³ per household, or 0.83 m³ per person per annum). Where it was necessary to convert fuel wood and pole units from cubic metres to tonnes, a factor of 0.8 was applied (following Odendaal et al., 1983).

Since products consumed directly by households have no market value, they were valued at replacement cost, i.e. using the market prices received when these products were sold. However, there is a dearth of current price information for informal market products, so the approach taken to value products was to estimate averages for prices recorded in surveys and in the literature from the southern African region. Prices for fuel wood were obtained from our own survey and from LaFranchi (1996), Loxton, Venn & Associates (1986), Ntshona (2002), and two case studies by Shackleton et al. (2002). Prices for poles were taken from our own study and that of Loxton, Venn & Associates (1986), and from two case studies by LaFranchi (1996) and three by Shackleton et al. (2002). Ex-sawmill prices for commercial saw timber (mostly blend prices for B. plurijuga and P. angolensis) were obtained from case studies in Namibia (Björkmann, 1999; Loxton, Venn & Associates & Plan Medewerkers, 1985) and Botswana (Norwegian Forestry Society, 1992). Several studies in southern Africa have provided estimates of the values of use of both NTFPs and fuel wood and we used an average of the ratio between the values for NTFP and fuel wood use. We estimated that for every N$1.00 of fuel wood value produced, N$0.64 of NTFPs was produced. We then applied this ratio to the fuel wood use values for the Regions to get a measure of value for NTFP use.

2.5 Potential use of woody resources

Forest asset account valuation requires a measure of the potential of the standing forest stocks, as measured in the physical asset account, to produce resource rent in the future. Depending on the forest product itself, only part of the total standing stock is physically suitable for use. Furthermore, at least in Namibia, owing to government policy and economic viability, only part of the physically suitable stock will be exploited in the future.

The physical asset account includes estimates of physically usable volumes for high-quality and medium-quality commercial saw timber and these volumes make up the physical saw timber potential. The proportions of standing woody volume that were physically usable for fuel wood and poles were estimated at 90 per cent and 15 per cent, respectively (using estimates from – in order of importance – Kanime & Laamanen, 2003; Mwilima & Boois, 2004; Boois, 2004) and applied to the standing volumes to derive potential volumes and yields.

It should be noted that there is some overlap between the physically suitable potential for different uses. Thus, for example, should all the physical fuel wood potential be used, then some pole potential would have to be sacrificed. A choice would then be required, and normally the more valuable use would take priority.

Mean annual increments (MAIs) were estimated to gauge the sustainable annual production potential of physically usable standing volumes for fuel wood, poles and saw timber. Measurements of MAI were not made during the national forest inventory; rather, we used the few available estimates from comparable habitats around the southern African region from, in order of importance: Nickerson (1984), ERL Energy Resources Ltd (1985), Loxton, Venn & Associates (1986), Musokotwane & Kufakwandi (1986), Loxton, Venn & Associates & Plan Medewerkers (1985), Millington et al. (1989), Timberlake (1999) and Scholes & Walker (1993).

The main constraints to economically viable fuel wood, pole and NTFP harvesting are remoteness, transport costs, spatially restricted markets, and market size. We assumed that the use of these products would grow as human populations grew, and as new initiatives developed in the harvesting and sale of these products for commercial purposes. In those Regions where physically suitable potential existed and within a 30-year future discount period, fuel wood harvesting activities were calculated to increase by 200 per cent to account for anticipated rural population growth, and by a further 150 per cent to account for new initiatives in commercial harvesting for urban markets. Similarly, pole use could be expected to increase by 200 per cent as populations grew, and by a further 140 per cent as new commercial production initiatives developed. Harvest predictions for each Region were restricted so as not to exceed the estimated sustainable yields of physically suitable material in the Region concerned. It was assumed that demand in excess of future production in these Regions would be met from other non-timber sources, even though some of the demand may be met from harvests in other more productive Regions.

No current saw timber harvesting takes place in Namibia at present, but it will probably be developed and expanded within the next 30 years. Potential extraction volumes – and yields – could be derived from the inventory volume data. Nevertheless, about 50 per cent of the saw timber resource was considered too remote and inaccessible for economically viable use in the next 30 years.

2.6 Valuation of flow and asset accounts

The value of the current and potential output of forest resources is the product of the volumes produced and the market prices, as explained above. A proportion of this output represents the direct contribution of the resource in terms of value-added to the GNP, as measured in the flow account. Another proportion of this output represents the amount of economic rent that the resource generates.

Use was made of a system of enterprise models, originally developed by Barnes (1998), to measure the financial and economic values associated with forest resource use. These are detailed budget and cost–benefit analyses that measure returns to investors as well as the national economy. Such models have been developed for harvesting for household fuel wood, poles, thatching grass, craft materials and so on, and harvesting for commercial timber and sawmilling (unpublished data, Directorate of Environmental Affairs, Ministry of Environment and Tourism, 2004).

For the monetary flow account, the enterprise models were used to calculate the direct contributions of forest uses in terms of value-added to the GNP. Value-added is defined as the return to the internal factors of production in the activity, namely capital, labour and entrepreneurship. The value-added was calculated in the models as a residual by subtracting intermediate expenditures from the gross output or total revenue. Value-added/output ratios were then applied in the flow account to determine the direct economic contribution of forest use activities. The direct contribution creates further demand in the broader economy through indirect multiplier and linkage effects. This represents the total economic contribution, or impact of forest use in the economy. As an example, using transport services in fuel wood harvesting would indirectly involve further value-added being generated in the transport sector. The flow account included a measure of this total impact. To measure the total economic contribution, a social accounting matrix model of the Namibian economy was used (Lange et al., 2004).

For the monetary asset account, the financial and economic models were also used to calculate the resource rents generated in forest use activities. These economic rents were also calculated as a residual – by subtracting intermediate expenditures, compensation of employees, consumption of fixed capital, and normal profit from the gross output or total revenue. Normal profit was assumed to be a 15 per cent return on initial fixed capital. The rent calculations were used in valuing the assets, using the Net Present Value Method. The portion of standing woody resource stocks that was not likely to be used economically in the ensuing 30 years was given a zero value in the monetary asset account.

Resource rents captured by government, through taxes, licence fees, etc., were also calculated using the models mentioned. Further details on rent capture by government were obtained from Directorate of Forestry records of sales of licences to harvest, transport and trade in forest products (unpublished data, Directorate of Forestry, Ministry of Environment and Tourism, 2004).

3. Results and discussion

3.1 Physical forest account

Table 2 shows the physical volume estimates for woody resources in Namibia. Total volume for Namibia is an estimated 257 million m³. Volumes per hectare area range from 21.37 m³ in the northeast (where volumes of 70–80 m³ per hectare are found locally in better developed teak woodlands) to 0.05 m³ in the southern Karas Region, where trees are stunted and restricted largely to drainage lines. The growth rates (MAIs) are averaged for the range of species in the woody community. As stated, there are no reliable data available to measure change in woody biomass due to the conversion of forest to cropland, or due to damage by fire or wild animals during 2004. Nonetheless, in the context of the total stock, such change is likely to be extremely small.

Table 2: Estimated forest standing volumes per hectare and aggregate, and physically suitable annual yields for fuel wood, poles and saw timber, 2004

Region	Forest biomass^a		Physically suitable annual yield^b
	Density	Volume	Fuel wood	Poles	Saw timber
	’000 m^3/ha	’000 m^3	’000 m^3/year	’000 m^3/year	’000 m^3/year
Caprivi	21.37	30 916.0	2504.2	417.4	118.8
Erongo	0.10	635.9	22.9	3.8	0.0
Hardap	0.10	1096.6	39.5	6.6	0.0
Karas	0.05	805.4	21.7	3.6	0.0
Kavango	18.00	87 269.4	6283.4	1047.2	48.2
Khomas	0.25	921.5	33.2	5.5	0.0
Kunene	0.20	2303.1	62.2	10.4	0.0
Ohangwena	20.00	21 388.0	1539.9	256.7	4.3
Omaheke	2.00	16 888.0	760.0	126.7	0.0
Omusati	3.22	8538.4	384.2	64.0	0.0
Oshana	0.90	781.4	35.2	5.9	0.0
Oshikoto	11.44	44 237.3	2388.8	398.1	0.0
Otjozondjupa	3.90	41 080.3	2218.3	369.7	0.0
Total		256 861.3	16 293.5	2715.6	171.3
^aTotal standing stock of woody resources.
^bPhysically suitable or usable yields from standing biomass, not necessarily economically exploitable.

Table 2 also shows the estimated sustainable harvest of fuel wood, poles and saw timber from the total stocks, in terms of what is physically usable or suitable for use. These yields are some 16.3 million m³ per annum for fuel wood, 2.7 million m³ per annum for poles and 0.17 million m³ per annum for saw timber.

3.2 The value of forest use

Table 3 shows the estimated volumes actually harvested in 2004 – some 1 million m³ of fuel wood and 334,000 m³ of poles. It also shows the estimated economically usable annual potential yields; after 30 years of expected future expansion, taking into account the economic constraints on the increase in use, the current harvest could more than triple. In the longer term, as increasing human populations and infrastructure allow more of the forest stocks to be economically harvested, use of the potential yields should grow even further. The currently used volumes (i.e. 6 per cent for fuel wood and 12 per cent for poles) are small in relation to the physically usable potential (see Table 2). Economically feasible use in the next 30 years is likely to take only a portion of the physically usable sustainable yield, i.e. some 16 per cent for fuel wood, 27 per cent for poles and 50 per cent for saw timber.

Table 3: Estimated harvest volumes for fuel wood, poles and saw timber, 2004, and their economically usable potential annual yield after 30 years

Region	Current annual harvest			Projected annual harvest in 30 years time^a
	Fuel wood	Poles	Saw timber	Fuel wood	Poles	Saw timber
	’000 m^3/year	’000 m^3/year	’000 m^3/year	’000 m^3/year	’000 m^3/year	’000 m^3/year
Caprivi	65.8	5.5	0.0	220.0	18.5	59.4
Erongo	21.8	6.8	0.0	22.9	3.8	0.0
Hardap	34.8	0.3	0.0	39.5	1.0	0.0
Karas	22.9	4.1	0.0	21.7	3.6	0.0
Kavango	141.4	52.5	0.0	472.9	175.7	24.1
Khomas	21.4	2.1	0.0	33.2	5.5	0.0
Kunene	47.6	2.4	0.0	62.2	7.9	0.0
Ohangwena	171.2	59.1	0.0	572.4	197.7	2.1
Omaheke	44.3	0.7	0.0	148.2	2.5	0.0
Omusati	175.0	77.4	0.0	384.2	64.0	0.0
Oshana	91.5	46.0	0.0	35.2	5.9	0.0
Oshikoto	114.2	74.1	0.0	381.8	247.6	0.0
Otjozondjupa	70.8	3.4	0.0	236.8	11.3	0.0
Total	1022.7	334.4	0.0	2631.0	745.0	85.6
^aExpected yields after economically feasible expansion over the next 30 years.

It is clear that at the national level the annual harvest, both current and expected in 30 years' time, is far below the physical potential. At a regional level, however, there may be depletion. Comparison between estimated current harvests (Table 3) and the estimated physically suitable sustainable yields (Table 2) suggests that current harvest might exceed sustainable yields in Karas and Oshana for fuel wood, and in Erongo, Karas, Omusati and Oshana for poles. This was first detected by Selänniemi et al. (2000a, 200b) in the forest inventory reports for the Omusati and Oshana Regions. Although demand in these Regions may be being met from harvests in adjacent Regions, local-level studies (Erkkilä, 2001) have shown evidence of over-harvesting in, for example, Omusati and Oshana, and elsewhere in the immediate vicinity of concentrated settlements. Solutions to regional and local depletion problems must be found through greater commercialisation of fuel wood and pole production in close concert with local forest management initiatives.

Table 4 shows some financial and economic characteristics of forest use enterprises, as extracted from the financial and economic models described above. Typical enterprises for small-scale fuel wood, pole and NTFP harvesting are depicted. The NTFP example used was grass collection. The saw timber extraction model is of a large-scale commercial enterprise producing sawmilled B. plurijuga and P. angolensis planks. It is assumed that enterprises make efficient use of capital and are financially profitable for the resource user or investor. The proportion of resource rent to output is high in all enterprises, notably in small-scale enterprises, particularly fuel wood harvesting. This is to be expected since labour inputs make up a significant proportion of costs.

Table 4: Characteristics of some typical forest use enterprises, extracted from empirically derived financial end economic models, 2004

Characteristic	Units	Enterprises
		Small-scale: fuel wood^a	Small-scale: Poles^a	Commercial: saw timber^b	Small-scale: NTFPs^c
Volume produced	m^3/annum	5.50	6.12	15 000	–
Financial (private) values
Initial capital	N$	900	900	6 059 900	1100
Gross output	N$/annum	4300	2100	23 003 300	3400
Variable costs	N$/annum	100	100	9 060 400	600
Fixed costs	N$/annum	500	400	4 513 100	1000
Net profit	N$/annum	3700	1500	9 429 800	1800
Internal rate of return	Per cent, 10 years	153	84	37	79
Net present value	N$, 10 years	19 300	7700	35 650 800	7800
Economic values
Gross value added^d	N$/annum	4300	1900	16 150 900	2200
Value added/output	Per cent	94	89	66	67
Resource rent^e	N$/annum	3600	1500	9 228 100	1800
Rent/output	Per cent	84	73	40	62
^aData from our own specific survey in the communal areas of the north-central Regions; some parameters differ from those used in the accounts (Table 1).
^bDerived from data in Bjorkmann (1999) in Namibia, and Norwegian Forestry Society (1992) in Botswana.
^cData from grass collection enterprises: derived from this study in the Caprivi Region, and from Terry (1999) in Botswana.
^dGross output, less intermediate expenditures.
^eGross output, less intermediate expenditures, cost of labour, and cost of capital.

Apart from basic sawmilling, the forest accounts do not include the processing of forest products. The use of forest products in building, furniture manufacturing and craft production is therefore excluded. Nonetheless, the value of these activities is captured in the estimates of the indirect contribution that forest use makes to the economy (see Table 5). It is interesting that enterprise models of craft production (e.g. Terry, 1999) appear to show relatively low generation of value-added and resource rent, compared with raw material collection enterprises like those shown in Table 4.

Table 5: Estimated gross output and direct and total contributions of use of fuel wood, poles, saw timber and non-timber forest products in the economy, 2004

Contribution	Fuel wood N$m	Poles N$m	Saw timber N$m	NTFPs N$m	Total N$m
Gross output in forest use sector
Total	648.3	176.7	0.0	415.7	1241.0
Direct contribution to GNP by Region^a
Caprivi	39.2	2.6	0.0	18.2	60.1
Erongo	12.9	3.2	0.0	6.0	22.1
Hardap	20.4	0.1	0.0	9.5	30.0
Karas	13.6	1.9	0.0	6.3	21.9
Kavango	84.3	24.7	0.0	39.1	148.2
Khomas	12.8	1.0	0.0	5.9	19.7
Kunene	28.4	1.1	0.0	13.2	42.7
Ohangwena	102.1	27.8	0.0	47.4	177.3
Omaheke	26.4	0.4	0.0	12.3	39.0
Omusati	104.4	36.5	0.0	48.4	189.2
Oshana	54.5	21.7	0.0	25.3	101.5
Oshikoto	68.1	34.9	0.0	31.6	134.5
Otjozondjupa	42.2	1.6	0.0	19.6	63.4
Total	609.3	157.5	0.0	282.8	1049.6
Total (both direct and indirect) contribution to GNP^b
Total	966.0	263.6	0.0	619.5	1849.0
^aDirect contribution of the forest use industry to the economy, in terms of value added to GNP.
^bTotal direct contribution to, and indirect impact on, the economy, in terms of value added to GNP, measured using a social accounting matrix of the Namibian economy (Lange et al., 2004).

Table 5 shows the estimated value of Namibia's use of forest resources in 2004. Total output in the forest use sector was N$1.2 billion. This sector contributed N$1.0 million of direct value-added to the GNP, and the total direct and indirect impact on the GNP was N$1.8 billion. The most significant component of forest use was fuel wood harvesting, followed by NTFP use and pole use. No legal saw-timber extraction took place at the time of the study.

The total direct value-added contribution of the forest use sector of N$1.0 billion represents approximately 3 per cent of GNP. This proportion can be compared with the estimated contributions made by other sectors (CBS, 2004): 4.6 per cent for agriculture, 5 per cent for fishing, 6.8 per cent for mining and 6 per cent for tourism (unpublished data, Directorate of Environmental Affairs, Ministry of Environment and Tourism, 2004). Much of the contribution of the forest use sector, as measured in the accounts described here (Table 5), is informal, and is not likely to have been fully captured in the national accounts to date. It thus largely represents an additional contribution.

In largely formal sectors such as fisheries and mining, it behoves the government to capture as much as possible of the resource rents (excess profits) generated, so that these can be redistributed within the economy and contribute to development. Yet forest use is currently largely informal, and much of the resource rent accrues to low-income rural households in communal areas. These poor households would themselves be targets for any national redistribution of rents captured from the forestry sector. Government should thus confine forest resource rent capture to commercial forest uses such as saw timber extraction. Currently, rent capture by government is largely limited to the Directorate of Forestry's collection of licence fees, since almost no taxes are paid. Licence fees tend to be concentrated on commercial forest use activities where trading is involved, and product sales constitute a very small proportion of output. Data to hand on licence fee collection is limited, but we roughly estimate that only some 0.5 per cent of resource rents are currently captured that way.

3.3 The value of Namibia's standing forest assets

Table 6 shows estimates of the value of the forest assets or standing stocks, the volumes of which are shown in Table 2. The values represent the ability of these stocks to generate resource rent in the future. As stated, these values were estimated using the Net Present Value Method. Forest stocks tend to be stable, unlike those of other natural resources such as marine fish or exploitable mineral reserves, and this makes it relatively easy to predict use patterns. Our estimates of likely economically viable growth in forest use over the next 30 years (described in Table 3) provided a good basis for applying the Net Present Value Method.

Table 6: The value of Namibia's standing forest stocks in 2004, estimated in terms of the resource rent that could be generated from them during the next 30 years, using the Net Present Value Method

Category of comparison	Value of standing forest resources
	Fuel wood N$m	Poles N$m	Saw timber N$m	NTFPs N$m	Total N$m
By Region at 6% discount
Caprivi	743	45	440	388	1616
Erongo	162	19	0	110	291
Hardap	272	2	0	185	460
Karas	150	18	0	102	271
Kavango	1597	430	178	834	3040
Khomas	203	17	0	138	358
Kunene	413	19	0	281	713
Ohangwena	1933	484	16	1010	3443
Omaheke	501	6	0	262	768
Omusati	1869	326	0	1271	3466
Oshana	270	43	0	184	497
Oshikoto	1289	606	0	674	2569
Otjozondjupa	800	28	0	418	1245
Total at 6%	10 202	2043	634	5857	18 737
Sensitivity to discount rate
Total at 2%	18 989	3789	1151	10 867	34 796
Total at 4%	13 689	2737	839	7846	25 110
Total at 6%	10 202	2043	634	5857	18 737
Total at 8%	7844	1577	494	4509	14 424
Total at 10%	6203	1252	396	3570	11 421
^aNet present value of future rents generated in expected growth over the next 30 years. The basic model used for net present values contains predictions about the future streams of rent and the discount rate.

Several discount rates were tested. We consider a future real discount rate of 6 per cent as the most likely, and so Table 5 shows net present value estimates at 6 per cent discount for the Regions and the whole country. The total value of the stocks is N$19 billion. If discount rates between 2 per cent and 10 per cent are used, the stock value varies between N$35 billion and N$11 billion.

Table 7 compares our forest asset values with the estimates made for some other Namibian natural resources. The figures in this table suggest that Namibia's forest assets have significant value, relative to those of other resources. One reason for the forests' high asset value is likely to be the fact that the (mostly informal) forest use activities generate relatively high resource rents per unit of output compared with other (formal) sectors such as fishing and mining. Like forest assets, fish, mineral, forest and wildlife assets are not incorporated into the national fixed capital stock accounts.

Table 7: Comparative estimates of asset value for some Namibian natural resources ,

Resource	Asset value (N$′000,000)
Fish^b	12 000
Minerals^c	14 300
Wildlife^d	10 600
Forests	18 700
Manufactured capital^e	82 000
Total national wealth^f	137 500
^aNet present values of future flows of resource rents from the stocks.
^bDerived from Lange (2004b).
^cDerived from Lange & Hassan (2003).
^dVery approximate (unpublished accounts, 2004; Directorate of Environmental Affairs, Ministry of Environment and Tourism).
^eFixed capital stock; derived from the national accounts (CBS, 2004); includes tangible, produced assets.
^fPartial estimate only; excludes, for example, some asset values for land, water and tourism.

4. Conclusion

The completion of the national forest inventory enabled the completion of a set of preliminary forest accounts for Namibia in 2004. These conform to the internationally recognised IEEA methodology. Volumes of woody resources per unit area range from 21.4 m³ per hectare in the northeast to 0.05 m³ in the south. The total woody resources volume for Namibia is an estimated 257 million m³. There are no reliable data available to measure change in woody biomass due to the conversion of forest to cropland or due to damage by fire or wild animals during 2004. However, such change is likely to be extremely small, annually, in the context of the total volume.

The currently used volumes are small in relation to the physically usable potential: 6 per cent for fuel wood and 12 per cent for poles. Similarly, economically feasible use in the next 30 years will probably make use of only a portion of the physically usable sustainable yields: some 16 per cent for fuel wood, 27 per cent for poles and 50 per cent for saw timber. It is clear that, while some regional and local forest resources adjacent to areas of dense human settlement are being depleted, this is not the case at the national level. Overall, stocks are not threatened, but local depletion problems will need to be solved through greater commercialisation of production and/or local forest management initiatives.

The value of current forest use in 2004, in terms of the gross output, was some N$1.2 billion. This made a direct contribution of N$1 billion to the GNP. The contribution amounted to some 3 per cent of the total GNP, comparing favourably with estimated proportions for other economic sectors based on natural resource use – 6.8 per cent for agriculture, 5 per cent for fishing, 6.8 per cent for mining and 6 per cent for tourism. Given the data previously available, it is highly unlikely that the national accounts have fully captured the current direct contribution of forest use, as measured here. In our study, the total direct and indirect economic impact of the forest use sector on the broader economy was estimated at N$1.8 billion.

Few taxes are paid, and government only captures a roughly estimated 0.5 per cent of the resource rents generated in the forest use sector through licence fees. Since most rents accrue to low-income rural households (which would normally be targeted for redistribution in any event) there is little reason for government to try to capture more.

Namibia's standing forest assets (the natural capital stock) were estimated to have a value of N$19 billion in 2004. Forest stocks represent a significant national asset, comparable with fish, minerals and wildlife. Fish, mineral, wildlife and forest stocks (as well as several other natural resources) are not included in the national account for manufactured capital stock, which was valued at some N$86 billion in 2004 and incorporated only owned or produced assets.

Some policy implications arise from the findings of this study. Attempts to ameliorate forest depletion need only be focused at regional and local levels, mostly in the immediate vicinity of denser human settlements. Here, initiatives to strengthen community management of forest resources and to promote commercial harvesting hold promise. Potential for commercial saw timber production also exists, and this should be developed through a comprehensive management plan to ensure sustainability. Capture of resource rent by government should be restricted largely to commercial forest use activities, especially any large-scale saw timber extraction.

Acknowledgements

The authors wish to thank their colleagues Glenn-Marie Lange and Jesper Stage for essential advice on forest accounting methods; the Directorate of Forestry, Ministry of Environment and Tourism, for inventory data and comments; Stig Johannson, Joseph Hailwa, Moses Chakanga, Sophia Swiegers, Michael Otsub for comments and Sandie Fitchat for editorial assistance; John Mendelsohn and Selma el Obeid, of the Research and Information Services of Namibia, and Esa Puustjarvi of Savcor Indufor, Finland, for comments and data; and the households, forest product traders and enumerators for data gathered in the field survey.