ARTICLE IN PRESS

Energy Policy 38 (2010) 3717–3730

Contents lists available at ScienceDirect

Energy Policy

0301-42

doi:10.1

n Tel.:

E-m

journal homepage: www.elsevier.com/locate/enpol

The forbidden fuel: Charcoal, urban woodfuel demand and supply dynamics,community forest management and woodfuel policy in Malawi

Leo Charles Zulu n

Michigan State University, Department of Geography, 103 Geography Building, East Lansing, MI 48823, USA

a r t i c l e i n f o

Article history:

Received 27 April 2009

Accepted 22 February 2010Available online 30 March 2010

Keywords:

Woodfuel

Community-based forest management

Malawi

15/$ - see front matter & 2010 Elsevier Ltd. A

016/j.enpol.2010.02.050

+1 517 432 4744.

ail address: zulu@msu.edu

a b s t r a c t

This article examines woodfuel policy challenges and opportunities in Malawi two decades after

woodfuel-crisis narratives and counter-narratives. A nuanced examination of woodfuel supply,

demand, use, and markets illuminated options to turn stagnant policies based on charcoal ‘bans’ and

fuel-substitution into proactive, realistic ones acknowledging woodfuel dominance and its socio-

economic importance. Findings revealed growing, spatially differentiated woodfuel deficits in southern

and central Malawi and around Blantyre, Zomba and Lilongwe cities. Poverty, limited electricity access,

reliability and generation exacerbated by tariff subsidies, and complex fuel-allocation decisions

restricted energy-ladder transitions from woodfuels to electricity, producing an enduring urban-energy

mix dominated by charcoal, thereby increasing wood consumption. Diverse socio-political interests

prevented lifting of the charcoal ‘ban’ despite progressive forest laws. Despite implementation

challenges, lessons already learnt, efficiency and poverty-reduction arguments, limited government

capacity, growing illegal production of charcoal in forest reserves, and its staying power, make targeted

community-based forest management (CBFM) approaches more practical for regulated, commercial

production of woodfuels than the status quo. New differentiated policies should include commercial

woodfuel production and licensing for revenue and ecological sustainability under CBFM or concessions

within and outside selected reserves, an enterprise-based approaches for poverty reduction,

smallholder/private tree-growing, woodfuel-energy conserving technologies, improved electricity

supply and agricultural productivity.

& 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Although ‘woodfuel crisis’ narratives of the 1970s and 1980sprojecting large deficits in sustainable woodfuel supply versusdemand had by the 1990s been dispelled as simplistic andalarmist, energy policies in many African countries still reflectthese misguided narratives, and sustainable woodfuel supplyremains a major and growing challenge. Counter-narratives fromseveral authors linked the crisis prognoses to misunderstandingsof energy-use patterns and responses to scarcity (e.g., Arnoldet al., 2006; Dewees, 1989, 1995a, b; Leach and Mearns, 1988).Misunderstandings included inflating fuelwood deficits throughlinear ‘woodfuel gap’ theory based on population, ignoringhousehold-fuel substitution with biomass residues or demand-management responses to scarcity; misdiagnosing economic forphysical scarcity, and overlooking significant supply from treesoutside forests; and misguided policies based on the erroneousnotion that woodfuel consumption is the main cause ofdeforestation instead of forest conversion to agriculture. Rather

ll rights reserved.

than build woodfuel/energy policies on such shaky ground, thecounter-narrative authors called for more sophisticated, consis-tent, evidence-driven, targeted, and realistic policies, drawing onlocal knowledge and participation.

There is inadequate evidence of a ‘return’ to national or globalwoodfuel crises; still, woodfuels have received less considerationthan is commensurate with their growing significance (Arnoldet al., 2006). The 1990s and 2000s saw a lull in donor interest andfunding of major woodfuel projects until a recent resurgencerelated to climate change.1 Although the beginning of the lullcoincided with the counter-narratives, it was more likely causedby diversion of aid from Africa to Eastern Europe and formerRussian republics after the end of the cold war, and by a new focuson global (rather than national) environmental issues followingthe Rio Summit, than by misinterpretation of the works ofcounter-narrative authors. Regardless, consumption of charcoal inAfrica is projected to double and firewood to increase by 24% from2000 to 2030 (Arnold et al., 2006). Localized or regional woodfuel

1 Interest re-emerged with debates and establishment in 2008 of a multi-

donor trust fund to finance the United Nations Reducing Emissions from

Deforestation and Forest Degradation in Developing Countries (REDD Program)

and start of carbon-sequestration projects.

ARTICLE IN PRESS

2 The BCFP ($16 million) started in 1986 as a woodfuel-crisis-era peri-urban

fuelwood project to supply Blantyre and Zomba cities. Failing to do so cost-

effectively, BCFP turned the forests to CBFM in 1996/1997 (Kalipeni and Zulu,

2002).

L.C. Zulu / Energy Policy 38 (2010) 3717–37303718

deficits, often hidden in national analyses, are expanding, e.g.,around major cities in Kenya, Mozambique, Tanzania, and Zambia(Mahiri and Howorth, 2001; Mwampamba, 2007; SEI, 2002).

So, how does Malawi, a predominantly rural country of 13.1million people heavily dependent on dwindling forest resourcesfor energy, one of the world’s poorest and most densely populatedcountries, meet growing energy needs in ecologically sustainableways while maximizing socio-economic benefits? That is acentury-old challenge that frames this study. National populationincreased 2.4-fold 1977–2008 and by 332% for Malawi’s four maincities 1987–2008 (NSO, 2008). Over 52% of the population livedbelow a poverty line of US $0.50 per person per day (GoM, 2007).Further, Malawi lost a significant share of its remaining forestcover from 45% of the 94,276 km2 land area in 1972 to 25.3% in1990 (Satellitbild, 1993), and an extra 25% (669,000 ha) by 2008as cultivated land expanded by 27% (MARGE, 2009).

Although Malawi depends on biomass for 89–93% of its energyneeds (GoM, 2003b; MARGE, 2009), the Department of Forestry(DoF) maintains a de facto ban, despite legalization of regulatedcommercial charcoal and firewood production under license fromindigenous wood under the largely progressive 1997 Forest Act,without providing real energy alternatives. The Department ofEnergy Affairs awaits energy transitions from woodfuels primarilyto electricity theorized under the energy ladder hypothesis as the

solution in a misguided and disjointed energy policy (Dewees,1995a, b; MARGE, 2009; Openshaw, 1997). This dominant energyhypothesis for developing countries predicts that householdstransition from ‘inferior,’ traditional biomass fuels to ‘modern’fuels (e.g., electricity, kerosene, and LPG) as incomes andurbanization increase (Arnold et al., 2006; Campbell et al., 2003;Hosier and Dowd, 1987; Leach, 1988). However, recent findingssuggest that factors other than income—energy preferences,convenience and insurance against unreliable supply—preventcomplete energy transitions and produce energy mixes instead(e.g., Hiemstra-van der Horst and Hovorka, 2008; Masera et al.,2000). Yet, the woodfuel ‘ban’ only criminalizes the largestforestry industry (woodfuels) worth MK15.5 billion (US$105 or3.5% of GDP in 2008) and the livelihoods of up to 200,000 peopleemployed full or part-time in the industry (MARGE, 2009) whileforgoing MK2.558 billion in potential VAT, without resolvingenergy or deforestation problems. Meanwhile, woodfuel produc-tion has become the political and public/media strawman for themore intractable, population-driven problem of forest clearancefor agricultural expansion (Dewees, 1995a; Kambewa et al., 2007).

The purpose of this article is to examine current woodfuelpolicy challenges and opportunities within Malawi’s politicallycharged deforestation debate, including potential for regulatedcommercial woodfuel production under community based forestmanagement (CBFM). I argue that (1) a nuanced understanding ofprinciple causes of deforestation, of woodfuel supply patterns andchains, woodfuel markets and trade, and patterns of energy use isessential to move a stagnant woodfuel policy based on fuel-substitution to a more proactive and realistic one acknowledgingwoodfuel dominance; (2) counter-indications to the energyladder hypothesis limit the potential for large-scale switchesfrom woodfuels to electricity under current conditions, therebymaintaining the dominance of woodfuels in the energy sector;and (3) community-based forest management is a more practicaloption for sustained commercial production of woodfuels thanthe status quo of prohibitive centralized control and an admin-istrative ban.

The article revisits the woodfuel situation in Malawi drawingprimarily on secondary data from recent woodfuel studies (e.g.,Arpaillange, 1996; Kambewa et al., 2007; MARGE, 2009; Open-shaw, 1997). It also draws upon 15 years of personal experience inMalawi forestry and fieldwork, including key-informant interviews,

focus groups, and survey of 381 households in 58 villages ofsouthern Malawi and forest inventory analysis of four communityforests in 2003. Recent developments prompted a revisit alsosuggested by Peter Dewees (1995a): adoption of new, CBFM-centered forest policy (GoM, 1996) and law (GoM, 1997), the 2003National Energy Policy, and the 2009 Malawi Biomass EnergyStrategy (BEST) study/report (MARGE, 2009), political decentraliza-tion in 1998; and changing patterns in woodfuel demand,supply, markets, trading, and use. CBFM devolves forest manage-ment responsibilities and rights to local communities for ecologi-cally sustainable utilization for poverty reduction. CBFM facessignificant challenges in Malawi (Blaikie, 2006; Jere et al., 2000;Kayambazinthu and Locke, 2002; Zulu, 2008, 2009a) and elsewhere(e.g., Campbell et al., 2001; Shackleton et al., 2002), but lessonshave also been learnt to improve it (e.g., Lund and Treue, 2008).

First the article reviews woodfuel policies and interventions inMalawi interrogates socio-political processes that underwritecurrent policy paralysis or the staying power of the charcoal ‘ban’despite being unrealistic and untenable. A revisit of woodfueldemand and supply confirms growing deficits in sustainablesupply in pockets around urban areas in central and southernMalawi, and surpluses in the north. Urban energy patternsrevealed growing woodfuel consumption and a shift mostly fromusing firewood to charcoal and electricity within an energy mix.Findings suggest a low transition ceiling or pace and a charcoal-dominated urban-energy landscape, given limited access, reliabilityand production capacity for growing demand, and unsustainabletariff subsidies. Decomposition of the value of woodfuels accruingto different participants along the supply chain revealed highestprofits to charcoal retailers and urban firewood wholesalers,significant bribes to transport illicit charcoal into town, henceroom for a formalized taxation system. CBFM experiences in theBlantyre City Fuelwood Project (BCFP) area in southern Malawiand elsewhere revealed challenges and opportunities for CBFM toregulate sustainable commercial woodfuel production in targetedbut dwindling areas with significant forest resources. Policydiscussions and recommendations precede the conclusion.

2. Background and methods

Malawi’s total per capita energy consumption is low (12.15 MJ,15.6% of upper-middle income countries) and dominated by thehousehold (84%) and biomass (93%) sectors (GoM, 2003b). Over91% of households used firewood and 6.2% charcoal for cooking(NSO, 2007). Most of the firewood and nearly all the charcoalcome from indigenous miombo woodlands, Malawi’s mainnatural vegetation type. Miombo are dominated by trees of theBrachystegia, Julbernardia, and Isoberlina genuses and cover2.7 million km2 in southern and eastern Africa (Campbell et al.,1996). Malawi’s miombo are usually in low rainfall areas(800 mm) supporting low productivity.

Community based forest management (CBFM) was formalizedunder the 1996 forest policy after a century of failure of top-downmanagement to arrest deforestation (see French, 1986). Politicaldecentralization reforms followed in 1998. The Norwegian fundedBlantyre City Fuelwood Project (BCFP) implemented the earliest(1997–2002) significant CBFM project involving 98 villagesmanaging 4700 ha of government-established Eucalyptus planta-tions and 15 villages managing portions of 12,000 ha of miombowoodlands in Blantyre, Chikwawa and Zomba districts.2 The

ARTICLE IN PRESS

Table 1Strategies to adapt to famine in southern Malawi, 2003.

Adaptation strategy Percent(n=381)

1. Reduced number of meals per day 48.0

2. Substituted maize with non-staple foods (e.g. pumpkins,

potatoes, wild foods)

45.9

3. Engage in piecework (ganyu) to earn income to buy food 39.1

4. Food grants from the government and other agencies 32.8

5. Produced or sold charcoal to buy maize 29.7

6. Sold livestock to buy food or exchanged livestock with food 16.8

7. Sold other crops (e.g. vegetables, cassava, potatoes, etc.), to

buy maize

16.0

8. No food deficit in year 14.2

9. Sold firewood to buy maize 11.8

10. Income from small businesses

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3719

Blantyre/Chikwawa component (over 800 km2) had suppliedcharcoal and firewood to Blantyre City since the 1930s. Rampantpoverty and low agricultural productivity increased livelihooddependency on forests mainly charcoal, from 25% of householdsin 1988 to 31% (up to 47% during famine), cf. 23% for subsistenceagriculture (Table 1). By 2003, 89% of respondents indicatedcharcoal and firewood sales and 76% poverty as the majorproximate and underlying causes of deforestation, respectively(Zulu, 2008).

The study used mixed social and natural science methods,relying primarily on secondary data from various reports andpublished work on woodfuel policies/laws, supply, demand,trading, and use patterns. Second, it drew on insights fromfieldwork conducted in 2003, observations during several visitsup to July 2009, and 18 years of personal experience in Malawiforestry. The fieldwork involved 381 household interviews, focusgroups with forest committees, charcoal traders and producers,and interviews with key government and community CBFM actorsin 58 villages in the Blantyre/Chikwawa portion of the BCFP. Itfocused on the socio-economic context, woodfuel supply-sidedynamics and day-to-day experiences in constructing CBFM. Thesurveyed households were selected by stepwise stratified randomand systematic sampling drawing 6–7 interviewees per selectedvillage. The study tangentially drew on forest-inventory analysis offour indigenous village forests to characterize forest productivityand sylvicultural viability of CBFM. Analysis involved descriptivestatistics, wood volume and growth-rate computations, synthesisand projections from existing data, and description.

3 Comparable interceptions were 1% Mozambique, 10% Zambia and 25%

Tanzania (SEI, 2002).4 The DoF impounded illegal charcoal/firewood and conveyance. Impounded

charcoal was sold for government revenue and the vehicles redeemable with fines

of MK15,000–MK30,000 (US$138–$276) per truck, depending on size

(US$1=MK108.57, September–December, 2003). A Blantyre transporter said he

could financially absorb one impounded charcoal trip in five.

3. Woodfuel policies, strategies and obstacles to regulatedproduction

Responsibility for Malawi’s woodfuel policies is splitmainly between the Department of Forestry (DoF) on productionand the Department of Energy Affairs addressing demandand broader energy issues. Both the National Forest Policy(1996) and National Energy Policy (2003) emphasize forest andenergy contributions to economic development and povertyreduction, and private-sector or community involvement. Themain goal of the latter is to ‘‘transform the country’s energyeconomy from one that is overly dependent on biomass to onewith a high modern energy component,’’ reduce dependence onbiomass from 93% in 2003 to 50% in 2020 and transition toalternatives, essentially electricity (GoM, 2003b). As for forests,controls largely continue British colonial policies dating back to

the Forest Ordinance of 1911, which focused on protectingforests from people through creation of forest reserves andrestricting commercial utilization rights to communities oncustomary lands.

DoF focus on protection and control of miombo forests andwide-ranging woodfuel interventions introduced by World Bank(WB) policy reviews and two Wood Energy Projects, WEPs(US $30.5 million, 1980–1993), have influenced postcolonialwoodfuel policy strategies significantly. WEP interventions in-cluded tree-harvesting restrictions, market-price interventions,smallholder/private tree growing, and energy conservationthrough improved woodfuel stoves and furnaces/boilers, andproduction of charcoal from waste plantation pinewood, butmany failed to meet expectations. The World Bank (1990b)reduced problems of deforestation and woodfuel deficits tomarket distortions and policy failures which led to suboptimalforest-resource use. It singled out uderpricing of government-grown wood and apparent flooding of the market with free/cheapwood from miombo forests with inadequate harvesting restric-tions for reducing prices and the true market value of wood,masking economic costs of deforestation, and discouraginginvestment in tree-growing and conservation. The DoF didnot fully implement WB-imposed annual 15% increases ingovernment-wood prices to reach cost recovery by 1996 (WorldBank, 1986). Dewees (1995a) argued that price interventionswould be ineffective regardless, because they applied only to the1% of national woodfuel demand supplied from governmentplantations/forests. Aggressive resource policing and licensingfailed to stem cheap woodfuel inflows into urban areas, whilehurting poor producers and traders disproportionally. ComparingDoF statistics on impounded charcoal sales for 2008 to demandestimates (MARGE, 2009) suggests that less than 0.02% of illegalcharcoal was intercepted.3 Observations in the BCFP area showedthat while wealthier transporters easily internalized charcoallosses to impoundment as operational costs, impoundment meantloss of capital or weeks of back-breaking work for poorerproducers or traders.4 Subsidized seedling production to promotesmallholder woodfuel production was misguided because amarket for poles provided adequate incentives, and a US $0.03cash bonus paid per planted tree surviving for 2 years proved toocostly to administer, without significantly improving tree growing(Dewees, 1995a). Regulated charcoal production from wasteindustrial pinewood in the Viphya DoF plantations (northernMalawi) and transportation to major urban areas (1986–1993)failed to alter the flow of and consumer preference for, indigenouscharcoal, or to penetrate the indigenous-charcoal marketingsystem, therefore failing to catch on (Habitat, 1993). In contrast,the stoves program was successful at dissemination wood-savingtechnologies (Arpaillange, 1996)

CBNRM policies and the Forest Act 1997 allow local commu-nities to participate in CBFM after demarcating a village forestarea, forming a community organization and electing a committeeto lead it, formulating forest rules, developing a forest manage-ment plan, and entering a forest management agreement (FMA)with the DoF (GoM, 1996, 1997, 2003a). The legislation grantscommunities new commercial rights to ‘‘develop a licensingsystem in collaboration with the Director of Forestry,’’ issuepermits, and set and charge fees for harvesting indigenous forest

ARTICLE IN PRESS

7 A World Bank study using remotely sensed data puts biomass estimates in

2004 at 388 million m3, having decreased 16% from 1990 (Bandopadhhyay et al.,

L.C. Zulu / Energy Policy 38 (2010) 3717–37303720

resources on unallocated customary lands managed under anFMA. However, commercial charcoal production, harvestingprotected tree species (now 14), and wood for processingindustries require a special license issued only by the Directorof Forestry (GoM, 1997). Requiring an applicant to prove that thewood is from a sustainable source with a management planessentially continues an administrative ban on miombo licensingdating back to the early 1990s, and criminalizes virtually allindigenous woodfuel production, transportation and trading.5

Thus, incentives for miombo-based CBFM and co-managementwere largely limited to minor or non-wood products. Severalfactors, including DoF responses to WB/WEP stipulations or policyrecommendations (July/August 1990), help explain persistence ofthe ‘ban’.

Social considerations, turf protection, DoF mistrust of commu-nity abilities and private sector intentions, vested politicalinterests, the negative image of the charcoal trade, and civilsociety/media pressure helped to maintain the charcoal ban. First,the DoF invoked affordability for the urban poor to justifyimplementing fewer wood-price increases than required (Dewees,1995a), and rejecting a WB recommendation to reduce invest-ment in ‘‘expensive peri-urban woodfuel plantations’’ (WorldBank, 1990b).6 Second was the ‘‘natural desire to protect theinterests and power of the DoF (and its contacts with privatesector forestry and powerful politicians)’’ (Kayambazinthu andLocke, 2002, p. 5) combined with community distrust (Zulu,2009a). The DoF also rejected turning ‘costly’ forest reserves overto community or private management because ‘‘the privateenterprise would overcut (clearfell)’’ and undermine water-catchment protection (World Bank, 1990b).

Third, vested interests by senior civil servants and politiciansinvolved in charcoal/timber trading and transportation favoredthe status quo. Even the minister responsible for forestry was in2003 implicated in illicit charcoal transportation. Politicians oftenabused their political power to reclaim impounded produce ortrucks. In one case, a DoF clampdown on rail trafficking ofcharcoal into Blantyre enforcing a parliamentary resolution in2002 ended prematurely on ‘instructions from above’, which alsopermitted certain influential transporters to transport charcoal.Politicians often supported illegal charcoal production for politicalgain. Enforcement patrols were perverted into cash generators forDoF-staff allowances and other extra-budgetary activities (creat-ing opportunities for abuse), or into tokenism ordered at will bypoliticians or senior staff to weather criticism of inaction. In 2002/2003 the Southern Region Forestry Office generated MK1.7million ($18,904) from 7550 bags of seized charcoal and fineson 51 trucks.

Fourth, charcoal has a negative reputation as a seedy andruthless business. Thus, a concession agreement allowing pinecharcoal production in the Viphya Plantations took many inter-ministerial discussions to overcome ‘‘the bad reputation whichcharcoal production on government plantations could have’’(World Bank, 1990a, p. 52). Fifth, while ostensibly favoring CBFM,civil society balked at deforestation, which they often reduced towoodfuel extraction, and demanded stringent top-down enforce-ment of restrictions. Finally, intra-government contradictionsallowed some government agencies to benefit from the illicitcharcoal trade. Local governments collected market for charcoalwhile the Malawi Revenue Authority collected taxes (MK20 perbag) for charcoal transported by train from Nsanje (Michael-Phiri,2006).

5 By July 2009, no charcoal license had been issued. Most DoF reserves/

plantations had no management plans.6 The DoF responded to the World Bank mission Aide Memoire in a letter

dated 17 August 1990.

4. Fuelwood crisis revisited – woodfuel supply and demand

Estimates of whether Malawi was in a woodfuel deficit orsurplus varied within and across scales. Relatively reliable supplyestimates still depend on projections from a 1993 forest-coverand biomass survey (Satellitbild, 1993), and put national stand-ing-wood stocks for 2008 at 400 million m3 and sustainablesupply at 29.8 million m3 yr�1—twice the annual demand of14.9 million m3 yr�1 (MARGE, 2009; see Table 2a).7 ‘Forests’ (atleast 20% tree cover) covered 2,248,500 ha or 24% of land area.Woodfuels constituted 91.6% of total wood demand (78.2%firewood, 13.4% charcoal), and poles and sawn wood 8.4%.However, inaccessibility of many forests to demand centers andcontinued depletion meant economic availability was lower thanphysical availability. Over half the forests were in protectedareas—88 forest reserves, 5 parks and 4 wildlife reserves(Table 2b). Girdis and Hoskote (2005) estimated 30 million m3

in sustainable annual yield for 1996. Orr et al. (1998) estimated25.8 million m3 for 1990, to be exceeded by demand by 2003,while the DoF declared ‘‘a chronic supply and demand imbalance’’of 7–8 against sustainable supplies of 15 million m3 in 2001(GoM, 2001). While varied methods and data explain some of thedisparities, the estimates still converged on growing wooddeficits.8

There was also considerable geographic variation. Reliable2008 estimates establish an 11% woodfuel deficit versus supplyfor the densely populated Southern Region, and small (5%)and large (77%) surpluses in the Central Region and sparselypopulated Northern Region, respectively (MARGE, 2009). Addi-tional wood from unsustainable land clearing for agriculturereduced demand to 89%, 79% and 21% of supply in the south,center and north, respectively (Table 2a). The Blantyre/Zomba andLilongwe urban woodfuel catchments had deficits of 25% (3% withwood from land clearings) and 31% (1% with clearings), respec-tively (Table 3). Incorporating biomass residue-use reduces thedeficits further. The Lilongwe City deficit grew 30% over 1996estimates (Openshaw, 1997), while that for Blantyre/Zombadropped 5%, probably due to data differences. Still, woodfuel-source areas expanded from 10 to 20 km in the 1980s to20–70 km in the 1990s, up to 150 km by 2006 (e.g., Balaka orNsanje to Blantyre).

Trees outside forests were important for woodfuel supply,contributing 40% of urban wood sources in 1996 (Openshaw,1997).9 Nationally, 54.6% of rural households in 2007 collectedtheir wood from unfarmed areas, 15.1% from forest reserves, 9.3%from community forests, and 4.1% from own woodlots/woodlands(NSO, 2007). Interpretation of analysis by Orr et al. (1998, p. 31,Fig. 8) of wood demand versus supply, which projected (1995)wood demand and (1992) supply estimates backwards to 1945and forward to 2005, suggests that wood from outside forestsexceeded within-forest supply around 1975/1976 and totalnational demand exceeded total sustainable supply from forestand non-forest sources around 2002. In contrast, there was a shiftin urban-charcoal supply from customary lands to protectedareas. By 2007 nearly 60% of charcoal wood for the four majorcities came from protected areas, 48% forest reserves and 10%national parks (Kambewa et al., 2007) compared to 12% (25%

2006).8 Common exclusion of branchwood and trees outside forests in DoF surveys

and vested professional interests in crisis narratives help to explain the 3–4-fold

lower supply estimates.9 ‘Counter-narrative authors’ also drew attention to trees outside forests (e.g.

Dewees, 1995b; Leach and Mearns, 1988).

ARTICLE IN PRESS

Table 2aForest cover and biomass and cover loss in Malawi, 1991–2008.

Region Population and density,

2008: Persons

(percent of total)

[persons/km2]a

Forest extent and

biomass, 2008b

Forest extent, 1991 Forest Change,

1991–2008c

Biomass Demand

versus Available

Supply, 2008d

1000 ha forest1 (%) million m3 (%) [forest

million m3]{m3/ha/yr}

1000 ha forest1 (%) Mean change 1000

ha (%) [percent/yr]

demand/supply1

(demand/supply2)

[demand/supply3]

1000 ha forest2 (%) 1000 ha forest2 (%)

North 1,698,502 (63) [45%] 868.2 (43.7%) 196.83 (48.4%)

[103.37]{4.50}

1124.3 (42.3%) �256.1 (�22.8%)

[�1.3% p.a.]

23% (21%) [21%]

2204.8 (48.1%) 2246.9 (38.6%)

Center 5,491,034 (154) [42%] 523.4 (26.3 119.70 (29.5%) [56.24]{4.48}

[

737.4 (27.7%) �214.0 (�19.0%)

[�1.1% p.a.]

95% (79%) [69%]

%)1294.6 (28.3%) 1529.4 (26.2%)

South 5,876,784 (185) [13%] 596.9 (30.0%) 89.79 (22.1%) [43.74]{4.57} 795.8 (28.0%) �198.9 (�17.7%)

[�1.0% p.a.]

111% (89%) [88%]

1082.4 (23.6%) 2053.4 (35.2%)

Total 13,066,320 (139)

[100%]

1988.5 406.380(100%)

[203.35]{4.52}

2657.5 (100%) �669.0 (�59.5%)

[�3.5% p.a.]

72% (62%) [59%]

(4581.8) [100%] 5829.7(100%)

a Source: 2008 population and housing census preliminary report (NSO, 2008), /http://www.nso.malawi.netS.b Forest1 represents closed forest and forest 2 includes forest in extensive agriculture areas (90% forest). Biomass includes all above-ground wood sources and percent,

and only the closed forest biomass and yield (m3/ha/yr). Source: Malawi Biomass Energy Strategy 2009 (MARGE, 2009) based on survey by Satellitbild (1993).c Forest change is based on only the closed forest class; forest1 in the third and fifth columns.d Supply1 is sustainable annual wood growth only, supply two is annual growth and wood from cleared land, supply3 is annual growth and residues (source: MARGE,

2009, p. 73).

Table 2bDistribution of forest resource by type.

Forest resourcea Hectares Percent

of forest

Forests on customary landb 1,988,255 47.80

Forest reserves (indigenous woodland) 828,451 19.92

National parks and wildlife reserves 1,100,151 26.45

Proposed reserves (indigenous woodland 137,726 3.31

Government plantations

Industrial plantations 75,000 1.80

Woodlots 15,000 0.36

Private sector

Fuelwood plantations 13,000 0.31

Industrial plantations 2000 0.05

Total 4,159,583 100.00

a Sources: Orr et al. (1998), GoM (2002) and Kainja (2000).b This is a mixed class of mostly agriculture with scattered trees.

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3721

for firewood) in 1996 when 63% came from customary lands(Openshaw, 1997). Thus, charcoal production appears a biggerthreat to protected than customary-land forests. The tobaccoindustry used 522,500 m3 (3.5% of 2008 wood demand), alocalized threat because tobacco depended on customary-landforests for 60% of their wood needs (GoM, 2001). Land clearancefor agriculture remained the largest cause of deforestation,compounded by low maize productivity and rapid expansion ofspecial crops. Between 1983 and 2007 maize productivity grew byonly 10%, population by 129% and cultivated land area by 112%(MARGE, 2009). Some 45,000 ha of forest and grassland wereconverted into agriculture in 2008. Based on these supply/demand patterns, interventions should focus on deficit areas inthe center and south, the Blantyre/Zomba and Lilongwe woodfuelcatchments, and tobacco producing areas.

10 I split MARGE’s production costs into 75% labor and 25% producer markup

based on observation that 75% of production costs were labor (MARGE, 2009, p. 49).

5. Urban woodfuel markets and woodfuel pricedecomposition

The woodfuel market is quite organized despite its illegality. Itinvolves multiple actors including rural wood sellers andwoodfuel producers, transporters, wholesalers, and retailers.

Interviews in the Blantyre catchment showed that commercialwood was rarely collected free of charge, contrary to commonbelief. Entry into the trade remained easy, requiring considerablelabor but little capital outlay, while DoF law enforcement wasvirtually non-existent.

The weighted average price of charcoal in 2008 for the fourmajor cities of Blantyre, Lilongwe, Mzuzu and Zomba and smallertowns was MK28.42/kg (range MK26.40 Mzuzu to MK33.30Lilongwe) and for firewood MK8.40/kg–MK6.60 in small towns,MK7.25 Zomba to MK8.73 Lilongwe (MARGE, 2009). Supply chainanalysis of the price build-up showed that transportation was thebiggest cost, but more critically so for firewood (63% of total costsand 38% of retail value versus 40% costs and 19% retail value forcharcoal), followed by production costs.10 Retailers enjoyed thehighest price markup over costs—43.5% for charcoal and 35.1% forfirewood—and the largest share of total markup per tonne (58.4%of MK14,748/tonne for charcoal and 65.0% of MK3355/tonne forfirewood), followed by roadside sellers and producers. Charcoaland firewood producers had 16.4% and 10.4% markups, and 12.2%and 8.9% or retail value, respectively. Wholesalers had the lowestmarkups—8 times lower than retailers for charcoal—but thehighest (transportation) costs (Fig. 1 and Table 4). The bulkier,lower-value firewood was much less profitable by weight.

However, even with 338 large-scale producers accounting for athird of production for the four main cities (Kambewa et al.,2007), the Malawian charcoal trade was not as dominated byurban elites as in other countries, e.g., Senegal (Larson and Ribot,2007; Ribot, 2009). In 2008 nearly 37,000 full-time jobs werecreated in rural areas producing and trading MK3258 million ($22million) in woodfuels (MARGE, 2009). In terms of the share offinal retail value for charcoal, estimates derived from the MARGE(2009) report were close to estimates by Kambewa et al. (2007)for Blantyre and Lilongwe: 25% accruing to producers versus27–33% (including packaging) in the Kambewa study, 19% totransportation versus 20–25%, and 30% to retailers versus 24–33%.While masking actual and relative share of profits, the Kambewastudy included ‘marketing’ fees amounting to 12–20% and 3% ofcharcoal-retail value accruing to bribes for forestry and police

ARTICLE IN PRESS

Table 3Urban annual firewood and charcoal consumption (1988, 1994, 2008) and main wood sources.

Urban area/

catchment,

population

2008

Charcoal, 2008a

tons major city

(1000 m3 city)

[tons all urban]

{1000 m3 all

urban)

Firewood, 2008a

tons major city

(1000 m3 city)

[tons all urban]

{1000 m3 all

urban)

Firewood,

1994b tons

(1000 m3

wood) [%

of total]

Charcoal,

1994c tons

(1000 m3

wood) [%

of total]

Charcoal,

1988d

tons (% of

total)

Sustainable wood supply,

2008e min. mean annual

increment, 1000 m3 (demand/

supply1) [demand/supply2]

{min. stock, million m3)

Main source areas for

firewood and charcoalf

Blantyre

661,444

78,143 (521.214)

[110,138]

{734.621}

161,214 (241.821)

[227,221]

{340.832}

170.25

(255.38)

[32.1%]

87,910

(558.23)

[69.5%]

30,000

(60.0%)

6228 (125%) [103%] {57.24} Chigumula, Zomba, Michiru FRs,

Neno/Mwanza area. Smaller

amounts from Mulanje, W.

Blantyre/Chikwawa and Zomba.

Main supply area radius is 80 km,

but over 150 km for train transport

(Makhanga, Balaka).

Lilongwe

669,021

51,481 (343.379)

[65,278] {435.406}

219,519 (329.279)

[278,351]

{417.576}

300.35

(450.52)

[56.6%]

32,970

(209.36])

[26.1%]

15,000

(30.0%)

5304 (131%) [101% {56.92 } Bunda FR, Dzalanyama FR, Dedza,

Nathenje for firewood; Salima,

Thuma and Dzalanyama FRs for

charcoal. Distances up to 80 km.

Mzuzu

128,432

6677 (44.537)

[12,217] {81.485}

48,439 (72.659)

[88,625] {132,93}

35.06

(52.59)

[6.6%]

4301

(27.31)

[3.4%]

2000

(4.0%)

7253 (22%) [21%] {113.20 } Chikangawa and Kaning’ina FRs;

Choma, Ekwendeni, Lusangazi,

Nkhata Bay, along Rumphi Road.

Shortest distances as low as 10 km.

Zomba

87,366

10,321 (68,844)

All urban

combined with

Blantyre

21,294 (31.941)

All urban

combined with

Blantyre

24.07

(37.30)

[4.7%]

1250

(7.94)

[1.0%]

3000

(6.0)

Combined with Blantyre

catchment

Zomba Mt. FR; charcoal Malosa/

Domasi area and Chingale. Longest

distance about 25 km.

1,546,263

[2,015,5-

67]

146,623 (977.974)

[189,504]

{1263.989}

450,466 (675.699)

[590,400]

{885.600}

530.53

(795.79)

[100%]

101,227

[802.84]

(100%)

50,000

(100%)

18,785 (87%) [84%] {227.35}

a Only household energy for the major city, and then all urban areas in the region of the major city. Derived from BEST per capita consumption estimates (MARGE,

2009) and 2008 population data (NSO, 2008).b Source: Openshaw (1997, Table 2.3). Unit conversion: 6.35 m3 of wood make 1 tonne of charcoal.c Source: Openshaw (1997). Unit conversion: 1.5 m3 of wood make 1 tonne of air dried wood.d Source: IPC, 1998 cited in World Bank (1990a).e Source: MARGE (2009). Supply1 is sustainably available woody biomass in the catchment area; Supply2 includes woody unsustainable agricultural clearance. Blantyre

and Zomba share a catchment area—results reported in Blantyre row.f Source: Kambewa et al. (2007), Michael-Phiri (2006), Openshaw (1997); and observations. FR stands for Forest Reserve.

Fig. 1. Decomposition of mean urban retail price, costs and markups for firewood (A) and charcoal (B) in Malawi Kwacha per tonne, 2008. aMonetary units are in MK100,

data are from Malawi Biomass Energy Survey (MARGE, 2009). bMonetary units are in MK200, data are from Malawi Biomass Energy Survey (MARGE, 2009).

L.C. Zulu / Energy Policy 38 (2010) 3717–37303722

officials for illicit passage into cities, and local government-market fees, respectively. The higher markup by retailers maypartly reflect these ‘marketing costs’ and higher profits fromselling in small amounts. However, multiple transport modes,routes and traders, and unknown proportions of firewood orcharcoal traversing them, and/or being intercepted by rent-seeking officials render further conjecture unproductive.

The main point is that the 12–20% of charcoal retail-valuegoing to bribes was internalized as a tax, albeit to personalpockets of corrupt officials. While legalization and regulatedcommercial production and transportation of woodfuels isunlikely to eliminate such corruption, it would remove a majorincentive (illegality) for police and forestry officials to extortbribes, and instead create opportunities for replacing this ‘private

tax’ with formal one(s), without disrupting the woodfuel market.The Malawi Revenue Authority (MRA) already collected a taxof MK20/bag (in 2006) for charcoal transported by rail fromNsanje into Blantyre (Michael-Phiri, 2006). Local governmentmarkets also collected retail market fees on the illegal charcoa-l—also a mark of policy contradictions among governmentagencies. Legalization/regulation would regularize such practicesand raise more government revenue. In 2008 the DoF generatedMK1,355,285 from sale of confiscated charcoal, 6.3% of totalrevenues of MK21.66 million (cf. the MK8.44 billion charcoalindustry), amounting to only 48.2 tonnes at the mean urbanprice of MK28,140/tonne. Thus enforcement of the ban inter-cepted only 0.02% of the 221,290 tonnes of charcoal traded inurban areas.

ARTICLE IN PRESS

Table 4Decomposition of mean urban retail price of charcoal and firewood in Malawi Kwacha per tonne, 2008.

Location on supply chain Cost a Cumulative cost Markup Selling price Percent markup Markup as % of

total markup

CharcoalRaw wood 2790 2790 0 2790 0.0 0.0

Production site 5377 8167 1793 9960 16.4 12.2

Roadside 1230 9397 3330 14,520 29.8 22.6

Wholesale, town 4270 13,667 1010 19,800 5.4 6.8

Retail, town 0 13,667 8615 28,415 43.5 58.4

Total 13,667 14,748 28,415 108.0b 100.0

FirewoodRaw wood 975 975 0 975 0 0.0

Production site 900 1875 300 2175 10.5 8.9

Roadside 695 2570 610 3480 28.0 18.2

Wholesale, town 2470 5040 265 6215 7.6 7.9

Retail, town 0 5040 2180 8395 35.1 65.0

Total 5040 3355 8395 66.6b 100.0

a Weighted mean costs, markup and retail price information for Blantyre, Lilongwe, Mzuzu and Zomba cities were obtained from the Malawi Biomass Energy Strategy

report (MARGE, 2009). The study assumed no costs associated with retailing.b The total percent markup is the total markup divided by total costs; not the sum of the individual percent markups for traders along the supply chain.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

1975 1980 1985 1990 1995 2000 2005 2010

Perc

ent o

f hou

seho

lds

Firewood Charcoal Electricity Paraffin and other

Fig. 2. Changing patterns of urban energy use in Malawi, 1981–2007.

Data sources: Energy Studies Unit (ESU, 1983) for 1981 data, 1998 national census (NSO, 1998), 2004 Integrated Household Survey 2004–2005 (NSO, 2005), and 2007

Welfare Monitoring Survey (GoM, 2007).

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3723

Findings also suggest that improvements in road networks andwoodfuel transportation efficiency may mitigate the transporta-tion bottleneck and reduce prices. Regional forestry officialsunderscored the growing significance of charcoal transportationby rail from Nsanje, Chikwawa and Balaka into Blantyre. Railtransportation was MK80/bag versus MK120/bag by truck(Michael-Phiri, 2006). Transport options and impacts on woodfuelsupply require further analysis.

11 Actual demand estimates are lower than BEST estimates, which were based

on higher 2008 urban (by 15%) and national (by 4%) population projections than

2008 census results.

6. Urban energy transition or an enduring firewood, charcoal,electricity mix?

Despite a significant movement up the energy ladder the past30 years, the main outcome has been an energy mix consistingmainly of firewood, charcoal, and electricity, but dominated bybiomass. Some 66% of urban households used more than one fuelfor cooking in 2008 (MARGE, 2009). Urban households usingfirewood as the main cooking fuel halved from 78% in 1981 to 35%in 2007, and charcoal and electricity users tripled from 14% to 44%and 6% to 20%, respectively (Fig. 2). Despite methodological andtemporal differences across studies, findings show a consistent,sizable ‘energy-ladder’ shift from firewood to charcoal—the‘middle class fuel’—driven as hypothesized by scarcity and

increasing costs and possibly by modestly rising incomes. Inenergy terms, urban households used 19,076 TJ of energy in 2008,56% as firewood (cf. 66% in 1994), 33% charcoal (24% in 1994) and adoubling in electricity use from 4% in 1994 to 9% (MARGE, 2009).11

Several factors undercut expected large-scale energy-laddertransitions from biomass to electricity. They include rapidpopulation growth, poverty, wood-to-charcoal energy conversionlosses, and low capitalization limiting electricity access, relia-bility, and generation, capacity. The 332% growth in population ofthe four major cities from 1987 to 2008 (2.4-fold increasenationally) means more people using the main energy forms(NSO, 2008). The national dominance of rural household energy(80%) over urban (12%), and of firewood (87%) over charcoal (6.7%)or residues, 6.4% (MARGE, 2009) limits potential impact of urbanenergy switches, although rural charcoal use grew from 0.4% in1998 to 4.2% of households in 2007 (NSO, 1998, 2007).

Malawi remains very poor (ranked 164/177 globally), incomesare chronically low and relatively stagnant, with high disparities.From 1990 to 2007, per capita GDP grew only 0.3% annually to

ARTICLE IN PRESS

Fig. 3. aTrends in charcoal, firewood, electricity and paraffin prices, 1996–2008.

Data source: National Statistical Office of Malawi and the Malawi Biomass Energy Strategy report (MARGE, 2009), and personal communication with Keith Openshaw and

Matthew Owen.

14 In November 2009, the government approved half (36%) the electricity tariff

raise requested.15 Increasing the electricity tariff from current levels (2.65 US cents/kWh) to

L.C. Zulu / Energy Policy 38 (2010) 3717–37303724

US$250, while mean annual inflation of 30% cut into real incomes(UNICEF, 2009). The 2004–2005 Integrated Household Surveyindicates that 54.2% of Malawians lived below an annual povertyline of MK16,165 per person (US$0.42/day), including 25% ofurban dwellers (NSO, 2005).12 Only 13% of Malawi’s populationwas in wage/salary employment, earning mean annual incomes ofMK73,274 in urban areas and MK10,867 in rural areas. Althoughmean annual per capita expenditure increased by 38.1% fromMK18872 in 1998 to MK26058 in 2005—MK52,594 (US$494) inurban and MK22,454 rural areas, 56% on food—2005 expenditureswere in real terms only 53.6% relative to 1998. NSO statisticsrevealed that non-food expenditure increased by 6.2% annually(total 37%) between 1998 and 2004 (NSO, 2005) while charcoal,firewood and electricity prices increased from 40.5% to 53.0%annually. From 1996 to 2008 mean firewood, charcoal, electricityand paraffin price increased 5.3, 4.9, 6.5, and 16 times,respectively, although in real terms (1996 base) energy pricesstagnated or declined, except for paraffin (Fig. 3). Poor householdssuffered more from these hikes than wealthier ones. According toArpaillange (1996) the poorest urban households bought charcoalin small amounts and paid 60% more (MK2.76/kg in 2004) thanthe wealthiest group (MK1.72/kg), which often bought charcoal inbulk.13 Kambewa et al. (2007) found that energy constituted23.7% (MK2087.23) of monthly household expenditures for thelowest of five income groups, and only 7.9% (MK4348.59) for thehighest (see also Fig. 4).

The government thus subsidizes electricity tariffs, therebyreducing capitalization needed for equipment maintenance and

12 The mean annual exchange rate for US$1 was MK106.36 in 2004 and

MK33.77 in 1998.13 This difference is well known and consistent with evidence from elsewhere.

development and undermining the quantity and quality ofelectricity supply, while benefiting the relatively wealthy.14 Thesubsidy amounted to US$830 annually for an upper-incomehousehold, and $320 and $80 for medium- and low-income ones,respectively (MARGE, 2009). It made cooking with electricitysuperficially the cheapest option, followed by firewood, thencharcoal.15 MARGE (2009) projects a 39% increase in commercial

annual firewood demand from 890,000 tonnes in 2008 to 1.25million tonnes in 2013, a doubling of charcoal demand from305,000 to 606,000 tonnes, and quadrupling of electricity demandfor cooking from 241 to 910 GWh. MARGE argue that meeting theincreased electricity demand for cooking alone would require392 MW in new electricity generation (cf. 285 MW in 2008)at a cost of US$600 million, and that ESCOM (the parastatalelectricity company) connects 25,000 new households to the gridannually. In 2008 ESCOM connected 15,000 households andhad a backlog of 15,000! While the government can choose tomaintain the subsidy as an incentive to meet fuel-switchingand forest conservation goals, the subsidy is currently self-defeating and commercially unsustainable. As MARGE aptlyobserve, the rationale of allocating more than half Malawi’selectricity to cooking when electricity demand is projected toexceed supply by 2011 even with the extra 64 MW coming online

an approximation of long run marginal cost (LRMC) of 9 US cents/kWh (cf. 24, 28

and 29 US cents in Rwanda, Kenya and Uganda, respectively), firewood would be

the cheapest cooking fuel followed by charcoal; electricity would be 44% more

costly than charcoal, but mean electricity cooking costs/household would triple

from MK9545 to MK27,877 (MARGE, 2009), a bitter pill and socio-political shock

unless incomes improve significantly.

ARTICLE IN PRESS

0

500

1000

1500

2000

2500

3000

3500

Low Density Medium Density High Density Unplanned Shanty

Mon

thly

exp

endi

ture

(Mal

awi K

wac

ha, M

K)

Electricity Charcoal Firewood

Fig. 4. The emerging energy mix: monthly energy expenditures in Blantyre, Lilongwe, Mzuzu and Zomba cities, 2007.

Source of data: Kambewa et al., 2007.

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3725

soon (MARGE, 2009). Industrial productivity and economicdevelopment already suffer from chronic load-shedding andcostly interruptions, and woodfuels remain dominant, isproblematic. Limited access to the power grid (only 25–30% ofurban households were legally or illegally connected by 2008),high costs of electric stoves, and energy preferences determinedless by price and more by task (e.g., charcoal for grilling) orseason, would lower MARGE projections of 470,000 householdsusing electricity by 2023. Contrary to ‘energy ladder’ predictions,the range of energy sources increased with incomes (Arpaillange,1996; Kambewa et al., 2007).

Finally, given firewood-to-charcoal as the dominant energy-ladder transition, wood consumption and ecological costs generallyincreased, ceteris paribus, because conversion-inefficiency lossesoften exceeded expected energy-efficiency gains from climbingthe energy ladder (higher energy content), even with the fairlysuccessful improved-stoves program. Adoption of the adaptedKenyan Jiko ceramic stove with thermal efficiencies up to 30–35%increased 5.5-fold in 15 years (6% in 1990 to 33% in 1995). Aremarkable 72% of charcoal users (cf. 4% for firewood) used thestove for cooking, consuming 25% less charcoal (with a potential of35–50% savings) than those using traditional metal stoves(Arpaillange, 1996). While, energy conserving technologies, in-cluding improved stoves, tobacco-barns and industrial boilersshould be a major part of energy policies, lower (than stove)charcoal conversion rates even in the reasonable 20%’s rangecommon in Malawi (e.g., Makungwa, 1997), and variable use ratesand efficiencies of the technologies fall short of recoveringcharcoal-conversion losses.16

17 Some women also participated. Charcoalers had a mean age of 36, 9 years

below the local average.

7. Potential and challenges of woodfuel regulation underCBFM

Poverty, livelihood dependency on forests, continuingdeforestation, and growing woodfuel deficits in the Blantyre-citycatchment make BCFP experiences relevant for assessingwoodfuel-production potential, and underscore the need for CBFMto improve conservation, livelihoods, and governance. For instance,

16 Conversion rates of 10–12% are often cited, but 19–20% (wet) and 25% for

air-dried wood have been recorded in Kenya and Tanzania (MARGE, 2009).

nearly every able-bodied male produced charcoal in the area, butmost characterized themselves as reluctant charcoalers forced bypoverty into illicit charcoaling despite its high labor/time demands,physical and health toll, economic insecurities, and indignities ofphysical abuse and harassment by DoF and police officials.17

Despite considerable challenges, indications suggest thatCBFM had better potential for regulated woodfuel productionthan the status quo. First, CBFM offers the spatial and socialpenetration needed to monitor resource use and administersupply-side woodfuel licensing systems. Charcoaling is hardto hide, and someone in the village often knows who is producingcharcoal, where. Production took weeks to months, producedtelltale smoke, and involved unconcealable social networksincluding village or clan heads, individuals and committeesselling wood to outsiders, villagers renting houses to itinerantcharcoaling gangs, money lenders, bicycle renters and repairers,charcoal scouts/packers, independent or ‘contract’ local produ-cers, and traders. Second, payment of formal and informal‘royalties’ for wood and bribes to corrupt local officials revealedpotential for formal royalties. In fact one village (Kabuluzi,Blantyre) demonstrated that communities can devise and imple-ment appropriate licensing/taxation systems.18 Committee mem-bers collected MK500–MK1500 per charcoal kiln or MK10/bag ofcharcoal produced, and established two village roadblocks toenforce the locally devised system.

Third, many causes of CBFM failure were more about flawedprocess than concept, hence correctable.19 The biggest was lack ofdownward accountability of forest committees, itself a product offailure to balance the exercise of power among key CBFM actors.Most committees operated as corrupt, elite clubs monopolizingdecision-making and benefits. Of 399 responses, only 11.8%indicated that CBFM decision-making was collective. Less than20% thought forest committees represented community interests,and 40.6% said committee members monopolized benefits.Relative success was associated with strong, creative, respected/feared, and charismatic village heads who were able to balance

18 Traders reported paying MK30/bag or MK600–MK2000 in bribes at the

three local roadblocks.19 Failure involved unsustainable forest use and CBFM institutions (see Zulu,

2008).

ARTICLE IN PRESS

22 The plan was not implemented. The DoF now recommends leaving behind

mature trees covering 20–30% of harvested compartments for seed/regeneration

and timber (IFMSLP, 2007).23 MARGE (2009) reports 60–123 m3 ha�1 in mean standing wood for

deciduous miombo, while Frost (1996, p. 24) reports stocks as low as 14 m3 ha�1

in Malawi, and 117 m3 ha�1 in Zambia.24 Details of the changes in species and structural diversity are covered

L.C. Zulu / Energy Policy 38 (2010) 3717–37303726

power relations (Zulu, 2008).20 Instead of privileging electedcommittees, new strategies should develop good village leader-ship and empower villagers to promote participatory decision-making, transparency and accountability. Transparent, negotiated,and clear cost/benefit sharing mechanisms were also essential.

Sound record keeping and financial accountability was anindispensable part of accountability and fleeting CBFM success. Itwill be crucial for managing new licensing systems, and needsmore attention than accorded in the BCFP. Record keeping wasgenerally poor, creating fertile ground for committee corruption,reduced community benefits, eroded trust, and dampenedparticipation. Twenty-two (56%) of a sub-sample of 39 villageshad incomplete, poor, or no financial records. Necessary trainingon financial accounting included complex material and lackedfollow-ups. It should be commensurate with low education levels,emphasize a learning-by-doing approach, and include regularfollow-ups.21

Appropriate rules are the heart of CBFM, and their formation,scaling, and practical enforcement were critical for CBFM success.Low participation (18% of respondents) in rule formulation, andrule scaling beyond the village-level to the chieftainship levelundermined local relevance and rule-enforcement (Zulu, 2008).Clear and enforceable accountability measures should be includedin village rules and bylaws, including requirements for commit-tees to present accounts to the community regularly. Findingspointed to dual-rule systems combining simple, informalcustomary rules and sanctioning at village level, with formal,wider-level (e.g. village-group) bylaws that bestow local andextra-village legitimacy and legal protections (Zulu, 2009a).

Lack of clarity of CBFM roles and unrealistic expectationsneeded redress. For instance, most committees operated as the

implementers or full-time workers—many paying themselvessecretly—not as leaders, coordinators, or mobilizers, as intended.Over 41% of respondents (n=352) indicated committee membersor paid labor was the main source of labor (23% cited the village‘community’—Zulu, 2009a). Initial DoF efforts to elevate electedcommittees over traditional leaders instead of harnessing com-plementarities caused conflict and undermined CBFM given thecentral role of these leaders over local social—cultural relations.Voluntarism or altruism as the key motivation for CBFM was alsounrealistic. Communities should be encouraged to considermaterial compensation for committee members and other labor/service needs transparently, commensurate with village re-sources, in order to reduce incentives for corruption and enhancebenefits. Some villages hired forest guards; others provided food,nominal cash payments, or wood for forest labor. As village headKapalaliza noted, ‘‘no one can be motivated to work hard withoutpay. Committee members who appear motivated derive someunderhand benefit.’’ Commercial woodfuel production can helpcommunities to get beyond the baseline, communal-benefitincentives promoted in CBFM to household-level incomes, whichappeared essential to sustain participation, especially amongmen. While technical forest management was less challengingthan institution building, it will need more attention.

In 1993 the BCFP prepared a sound miombo management planthat could be adapted to other areas with significant forests,including reserves. The plan split the 12,000 ha miombo forestinto three zones: ecologically sensitive areas for total protection,semi-sensitive areas limited to selective harvesting, and autilization zone for charcoal production based on a 25-yearsylvicultural rotation (i.e., it would be split into 25 equal-area

20 Powers of village heads have to be moderated, too, so chiefs are not too

weak or overbearing.21 Nearly 83% of respondents (n=377) had no formal, or never completed

primary, education.

compartments and one clear-harvested annually, in rotation).22

Trained BCFP communities exhibited impressive levels of techni-cal Eucalyptus plantation skills, but miombo management skills,including basic resource-assessment and management-planning,were poor, as they also are in the DoF. Despite high training costs,the EU-funded Social Forestry Training and Extension Project(1997–2003) successfully trained 14 communities in participatoryforest resource assessment (PFRA) and miombo managementplanning. Guidelines have been developed for wider application.

However, generally low productivity from ecological limitationsand continuing degradation limited prospects for sustained com-mercial charcoal production in much of the BCFP area. Forestinventory analysis confirmed slow tree recovery. One forest(Chigunkha) produced only 3.45 m3 ha�1 of standing wood(5.6 cm mean breast-height diameter) after 11 years of protectionand recovery. One of the most intact and mature woodlands,Kabuluzi (620 ha) had 75.1 m3 ha�1 standing wood, but no morethan 10 stems per hectare of the 25–50 cm diameter trees preferredfor charcoal production (Zulu, 2009b).23 On a 25-year rotation,each 25-ha compartment would yield 1878 m3, or 281.5 tonnes(10053 28-kg bags) of charcoal worth MK7.513 million (US$53,471)or MK12,117/ha. Approximately 35% of total markup or 51% of totalretail value would remain locally for Blantyre in 2008 prices(US$1=MK140.51 in second quarter, 2008), supporting 67 smallproducers (150 bags/producer/year).

Although woodfuel extraction per se did not cause significantdeforestation and is potentially renewable unless followed byconversion to agriculture, as has been found in Mozambique,Tanzania and Zambia (SEI, 2002), and globally (Geist and Lambin,2002), cycles of woodfuel extraction had by 2003 changed thespecies composition and reduced forest structure largely intoscrublands under 1.5–3.0 m tall.24 Remote sensing analysis withLandsat satellite imagery showed that net deforestation in theBCFP area was only 1.7% between 1989 and 2002 while 44% of thearea had degraded (thinned) considerably (Zulu, 2009b).25 Villa-gers would have to wait 15 or more years for the forest to recoverbefore producing charcoal, a tall order, as one villager’s (Malekano)quip illustrates: ‘‘do you expect me to work today so I can enjoythe money when I am dead?’’ However, many community forestswere smaller and more degraded than Kabuluzi, underminingviability of sustained charcoal production.26 Here, value-addingbased on minor forest enterprises and a focus on forestrehabilitation and tree growing make more sense. So, CBFM innot the panacea, and charcoal production would be more viableonly for areas with significant, contiguous forest resources.

8. Policy implications and recommendations

If the charcoal ‘ban’ and woodfuel policy are clearly notworking, why do they persist? The status quo ignores the socio-economic importance of woodfuels as Malawi’s major energysource despite growing demand and without providing a realalternative, criminalizes a MK15.5 billion industry that employs

elsewhere (Zulu, 2009b).25 Some 9.2% of forest converted to non-forest, including agriculture; 5.6% of

non-forest area became forest, mainly in BCFP Eucalyptus plantation and some

forest recovery.26 From 15 ha in the BCFP, 1111 registered forests in Lilongwe ranged from 0.1

to 70.0 ha!

ARTICLE IN PRESS

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3727

up to 200,000 people full or part-time, denies the governmentpotential tax revenues for conservation and other priorities aspublic funding dwindles, and promotes corruption—all withoutreducing deforestation. Diverse interests helped to maintain theselective devolution of commercial woodfuel production rightsunder otherwise progressive forest laws. They included govern-ment social redistribution considerations in favor of the ‘urbanpoor;’ DoF self-preservation or fear of the ‘‘professional dis-empowering’’ effect of CBFM (Blaikie, 2006) and mistrust ofcommunities; vested interests of some senior governmentofficials and politicians; a bad image associated with the charcoalbusiness; and discordant support of civil-society elements for aclampdown on charcoal, combined with DoF resource constraintsand usual bureaucratic resistance to change.

Charcoal is the most political natural resource after land, andthe current policy paralysis (ban) allows access to a lucrativecharcoal business and political patronage for government andpolitical elites. Similar self-interests and lackluster DoF supportalso recently scuttled efforts to privatize forest plantations after 4years of donor-funded preparation (DFID, 2006), thereby main-taining low wood prices and privileged access to timber-sawingpermits by elites. Recent interviews (June–July, 2009) of seniorDoF officers revealed that no one was willing to be associatedwith the politically sensitive decision of publicly lifting thecharcoal ‘ban.’ Similar bans were revoked in Chad and Tanzaniaafter causing social upheaval, and two decades of a continuingban in Uganda has not replaced charcoal as the major urban fuel(MARGE, 2009).

Mitigating these vested political, economic, and professionalinterests to create space for realistic woodfuel policies is achallenge requiring bold action. In Uganda and Senegal vestedinterests have allowed forest bureaucracies to retain authorityover commercial forest/woodfuel rights—hence sidelining com-munities—despite permissive CBFM laws (Ribot, 2009; Ribot andAgrawal, 2006). The Malawi DoF has historically limited thestandard forestry principle of wood harvesting for sustained andenhanced regeneration and productivity to planted exotic species,and prescribed preservation for indigenous forests. The use of pilotprojects, such as the EU-funded Improved Forest Management forSustainable Livelihoods Project, IFMSLP (2006–2012), at leastdebunks the ‘taboo’ of tree-harvesting in forest reserves and canhopefully demonstrate sustainable forest harvesting under CBFM.The IFMSLP gives forest licensing rights to 12 reserves covering150,000 ha (16% of total) to neighboring communalities under co-management agreements between the DoF and organized com-munities based on agreed forest rules, a forest management planand an enterprise-development plan (IFMSLP, 2007).

This study has shown that large-scale energy switches toelectricity and other alternatives expected under the NationalEnergy Policy (2003) are unrealistic in the short-medium term,and woodfuels will still dominate the energy scene. Woodfueldemand will actually increase, and energy policies should reflectthis reality.27 Even wealthier, more urbanized African countriesfacing nothing like Malawi’s low urban incomes; limitations onelectricity access, quality and quantity of supply; and outpricedparaffin and LPG alternatives (Fig. 3), have not achieved the fullenergy transitions posited under energy-ladder theory. Instead, asthe case of Maun in Botswana illustrates, energy mixes emergefrom constrained energy transitions reflecting factors beyondincomes, including choice of certain fuels for specific uses or times,convenience, and insurance against unreliable energy supply(Hiemstra-van der Horst and Hovorka, 2008). The dominant

27 BEST projections put biomass dependence in 2020 at 83%, not 50% (MARGE,

2009).

transition is firewood to charcoal (Fig. 2) because charcoal is moreconvenient, cleaner (less smoke), easier to store without degrada-tion, has double the energy content of firewood and morespecialized uses. However, the switch leads to higher woodconsumption because energy losses in conversion from firewoodto charcoal generally outweigh end-use efficiency savings fromgoing one rung up the energy ladder. Every person convertingfrom firewood to charcoal used 0.63 m3 (43%) more wood annuallythan before despite 20–44% in end-use efficiency gains from thehigher energy content and improved-stove use, depending onwhether one used traditional or improved stove before or after,based on 2008 per capita consumption estimates (MARGE, 2009).

Given its dominance in the energy sector, more attention shouldbe given to efficient production and supply of woodfuels. CBFMoffers a very good opportunity not as the panacea, but as a majorcomponent of sustainable woodfuel production. Although mixedCBFM performance justifies caution and socioecological factorslimit the viability of commercial woodfuel production under CBFMto diminishing areas with significant forest resources, socio-economic, efficiency, conservation, and social justice arguments,and failure of centralized approaches support CBFM. Currently,there is no real alternative to CBFM. As Lund and Treue (2008) alsoargue, most CBFM challenges arise from implementation ratherthan conceptual problems that can be corrected. Lessons from theBCFP suggest that instituting transparent accountability anddecision-making processes, including required regular financialreporting to communities, mechanisms for equitable cost/benefitsharing, and developing leadership skills of village heads to balancethe exercise of power among CBFM actors could mitigate the majorCBFM problem of elite captures (see Zulu, 2008). A simple villagelicensing system tried by Kabuluzi Village in W. Blantyre showsfeasibility of community production-side licensing for common-access forests while the IFMSLP illustrated viability under forest-reserve co-management. Given recent shifts to forest reserves asmajor sources of illegal charcoal (Kambewa et al., 2007), andlimited DoF capacity to reverse them, permitting charcoal produc-tion in reserves under the IFMSLP would be an excellent entrypoint for proactive regulated charcoal production already providedfor in the 1997 Forest Act.28 The IFMSLP has a revenue-sharingformula: 60% to communities, 30% to government, and 10% to areserve-wide oversight association. Government tenure of forestreserves facilitates relative control over licensing and forestharvesting, while the relative security of supply allows viableintroduction of more energy-efficient kilns, e.g., the half-orangebrick kiln successfully tested in the Viphya plantations, to reducewood wastage. As for the Viphya Charcoal Project, supply-chainanalysis has illustrated that transportation is not as prohibitive afactor as previously believed, especially for charcoal. Conveyancepermits could tax wood from non-CBFM or non-regulated sourcesmore than CBFM/regulated sources (see also MARGE, 2009) toencourage conservation.

Though CBFM-dependent commercial woodfuel productionpolicies are relatively new, yet Malawi can learn from othercountries. Several African countries use a combination of CBFMagreements and concessions supplemented with checkpoints onmajor urban feeder roads and varied taxation systems, althoughlevels of commercial rights devolved to communities vary(Table 5). Organizing charcoal producers into some umbrellacharcoal association(s) as done in Zambia and Sudan couldstreamline licensing to associations, which would further licenseothers and regulate activities of their members (see Mugo and

28 Legally, IFMSLP communities can choose charcoal production but the DoF,

which approves management plans, has not encouraged them for reasons

discussed earlier.

ARTICLE IN PRESS

Table 5Comparison of commercial charcoal and woodfuel policies and strategies in selected African countries.

Country FAO forest cover

2005a

Commercial utilization rights Legislated

licensing

scale

CBFM commercial rights Summary of key policy elements or

lessons

1000 ha

(%)

ha/

capita

Malawi 3.402

(36.2)

0.3 Transport and production licenses/

permits

National,

district,

village

Licensing rights (licenses, permits). � De jure rights commercial rights not

fully translated into reality.

� Commercial licensing of indigenous

woodlands suspended since mid-1990s.

Mozambique 1962

(24.6%)

1.0 Licenses, concessions. Provincial None � License fees go into Agrarian

Development Fund for small-scale rural

development.

� Fees at checkpoints on main city feeder

roads.

� Testing taxation systems for producers.

Transportation fee per bag and

15% restocking fee.

Senegal 8673

(45%)

0.8 Production, storage and

transportation licenses, concessions,

professional permits, quotas.

Region, Rural

Councils

Licensing & concessioning rights;

communities keep 70% of fines,

sales.

� Govt. forestry taxes go to National

Forestry Fund for forest rehabilitation.

� Quotas based on forest management

plan.

� Commercialization banned in non-

managed areas.

Tanzania 35,257

(39.9%)

1.0 Licenses and permits Rural

councils,

central

government

Licensing rights; 100% revenue

retention.

� Has longer history of decentralized

village governance.

� Lessons on institutional building.

Zambia 45,452

(57.1%)

4.0 Licenses and permits, charcoal

conveyance fee (US$0.10/bag).

National,

district,

village

Licensing rights (licenses, permits). � Use charcoal cooperatives.

� Fees at checkpoints on main city feeder

roads

Sudan 67,546

(28.4%)

2.0 Concessions, tender to charcoal

association, taxes for traders

National Via association membership;

villagers can grow crops in

harvested forests for few years

� Government-run plantation wood

production.

� Producers organized into Sudan

Charcoal Association.

� Separate Sudan National Corporation

regulates charcoal production and

trading.

Uganda 3627

(18.4%)

0.1 Licenses District, sub-

district

councils

Licenses, permits � Commercialization banned from

Central Forest Reserves (all forests

100 ha or more), allowed in buffer

zones.

� Very little forest land for local

management.

Sources: FAO (2007), Lund (2007), Mugo and Ong (2006), SEI (2002), and Ribot (2009).

L.C. Zulu / Energy Policy 38 (2010) 3717–37303728

Ong, 2006; SEI, 2002). Mozambique and Senegal levy a forestrestocking fee or equivalent which goes into a national forestryfund. Malawi should reactivate the Forest Development andManagement Fund provided for in the Forest Act 1997 (Part IX). Asemi-autonomous unit dedicated to woodfuel regulation(transportation, licensing, trading, standards, and demand/supply assessment), as in Sudan, could enhance woodfuelsupply and regulation. Management agreements should includeproduction quotas for monitoring supply, as in Senegal (Post andSnel, 2003). Finally, two studies covering 15 Tanzanian villagesshowed that village-based taxation and regulation of commercialforest utilization (including charcoal) increased revenues andcovered forest management costs, with a surplus (18%) for villagedevelopment, controlled utilization effectively, achieved woodgrowth in excess of harvesting, and enhanced transparency andaccountability over village funds and decision-making (Lund,2007; Lund and Treue, 2008). However, safeguards are needed forthe poor because regulations and taxes made the poor worse off(also see Fisher, 2004).

Ultimately, CBFM should be combined with other interven-tions to meet urban/rural, household/industrial woodfuel demandand support rural livelihoods sustainably. They include enterprisebased rural tree-growing and conservation to support livelihoodsand harness the importance of trees outside forests; expansion ofthe relatively successful stoves program for urban households andindustrial barn/boiler energy conservation, and improvements inelectricity access, reliability and generation capacity, and graftminimization measures that could include collection of woodfueltaxes by the Malawi Revenue Authority. Improving agriculturalproductivity, e.g. through agroforestry and irrigation, is a majorpart of the long-term solution.

9. Conclusion

This article has argued that a holistic examination of supplyand demand-side woodfuel dynamics, including quantities,sources, production relations, geographic patterns, trading,

ARTICLE IN PRESS

L.C. Zulu / Energy Policy 38 (2010) 3717–3730 3729

demand and energy-use patterns provided a nuanced under-standing of the woodfuel situation and realistic policy options.While there is a partial shift in urban-energy use from firewood tocharcoal, expected large-scale transitions from woodfuels toelectricity are currently unlikely.

Transitions are constrained by low incomes and lack of or lowincome growth, complex fuel-allocation decisions by householdusers, and limited access, unreliable supply, and low generationcapacity for electricity given unsustainable electricity subsidies thatundercut capital improvements. Thus, urban woodfuel demandcontinued to rise rapidly relative to supply within an enduringenergy mix of firewood, charcoal and electricity, dominated bywoodfuels. Policies should therefore reflect the reality of a woodfuel-dominated energy sector rather than unrealistically wait for energyswitches while banning the major energy source, without providingreal alternatives. This article further argued that the commercialwoodfuel ‘ban’ is anachronistic, ill-informed by a crisis discoursethat erroneously reduces deforestation to fuel-wood extraction andby overdependence on energy-transition theory. It is thereforecounterproductive, and untenable. It should be removed in ameasured manner that produces more win–win situations reflectingthe economic importance of woodfuels and their ecologicallysustainable supply while fighting poverty.

The community participation and decentralization genie hasbeen out of the bottle since the mid-1990s, and as Blaikie (2006)notes, CBFM appears to be here to stay. It is time to use lessons thathave been learnt in Malawian and elsewhere to improve CBFM. Thearticle has argued for proactive regulation of commercial firewoodand charcoal production under CBFM not as a wholesale panacea,but carefully targeted to customary-land areas with adequate woodand market linkages and to selected forest reserves under co-management agreements with communities and concession ar-rangements with private producers. This offers more realisticpromise toward meeting woodfuel demand, contributing to rurallivelihoods, and ecologically sustainable utilization than the statusquo of policy paralysis. CBFM constitutes a realistic response torecent shifts in urban charcoal-wood supply from customary landsto reserves. However, instead of a one-size-fits-all approach, CBFMshould be part of a differentiated but coordinated energy policyintegrating supply and demand dynamics. There is need forpromoting smallholder and private tree growing and conservationgiven that mismanagement and growing demographic pressurespromote continual forest fragmentation and degradation, limitingareas of viable commercial charcoal production from indigenouswood under CBFM. Expansion of energy conserving stoves for urbanhouseholds and barn/boiler technologies for small industries, andimprovements in electricity access, reliability, and generation arealso important in the short-medium term. Future improvements inagricultural productivity are needed to undercut the real majorcause of deforestation (forest conversion to farmland) and increasingeconomic dependency on forests.

This study also illustrates challenges of implementing otherwiseprogressive forestry laws within a neo-patrimonial politicizedenvironment. It contributes to debates and praxis on deforestation,woodfuel supply/crisis, and energy policy options in the context ofpoverty and environmental and political decentralization reforms.However, more data and research are needed, including updatingthe national forest-cover and biomass assessment incorporatingtrees outside forests, understanding long-term household responsesto, and allocation of resources during, wood scarcity, and improvedmodeling of woodfuel demand scenarios.

References

Arnold, J.E.M., Koohlin, G., Persson, R., Shephered, G., 2006. Woodfuels, livelihoods,and policy interventions: changing perspectives. World Development 34 (3),596–611.

Arpaillange, J., 1996. Urban Household Survey: Demand Side Strategy. Ministry ofEnergy and Mining, Lilongwe, Malawi.

Bandopadhhyay, S., Shyamsundar, P., Baccini, A., 2006. Forests, biomass use andpoverty in Malawi. Draft background paper for the Malawi Poverty Assess-ment. World Bank Policy Research Working Paper 4068. The World Bank,Washington, DC.

Blaikie, P., 2006. Is small really beautiful? Community-based natural resourcemanagement in Malawi and Botswana. World Development 34 (11),1942–1957.

Campbell, B., Frost, P., Byron, N., 1996. Miombo woodlands and their use: overviewand key issues. In: Campbell, B. (Ed.), The Miombo in Transition: Woodlandsand Welfare in Africa. CIFOR, Bogor, Indonesia, pp. 1–5.

Campbell, B., Mandondo, A., Nemarundwe, N., Sithole, B., de Jong, W., Luckert, M.,Matose, F., 2001. Challenges to proponents of common property resourcesystems: despairing voices of the social forests of Zimbabwe. Society & NaturalResources 29 (4), 589–600.

Campbell, B.M., Vermeulen, S.J., Mangono, J.J., Mabugu, R., 2003. The energytransition in action: urban domestic fuel choices in a changing Zimbabwe.Energy Policy 31, 553–562.

Dewees, P.A., 1989. The woodfuel crisis reconsidered: observations on thedynamics of abundance and scarcity. World Development 17 (8), 1159–1172.

Dewees, P.A., 1995a. Forestry policy and woodfuel markets in Malawi. NaturalResources Forum 19 (2), 143–152.

Dewees, P.A., 1995b. Trees on farms in Malawi: private investment, public policy,and farmer choice. World Development 23 (7), 1097.

DFID, 2006. Project Completion Report: Malawi Forest Sector Support Program(MFSSP). Draft. DIFD, Lilongwe, Malawi.

ESU, 1983. Malawi Rural Energy Survey. Energy Studies Unit (ESU), Lilongwe,Malawi.

FAO, 2007. State of World’s Forests 2007. FAO, Rome.Fisher, M., 2004. Household welfare and forest dependence in Southem Malawi.

Environment and Development Economics 9, 135–154.French, D., 1986. Confronting an unsolvable problem: deforestation in Malawi.

World Development 14 (4), 531–540.Frost, P., 1996. The ecology of Miombo woodlands. In: Campbell, B. (Ed.), The

Miombo in Transition: Woodlands and Welfare in Africa. Center forInternational Forestry Research, Bagor, Indonesia, pp. 11–57.

Geist, H.M., Lambin, E.F., 2002. Proximate causes and underlying driving forces oftropical deforestation. BioScience 52 (2), 143–150.

Girdis, D., Hoskote, M., 2005. Malawi: Rural Energy and Institutional Development.Energy Sector Management Assistance Program, ESMAP. International Bank forReconstruction, Washington, DC.

GoM, 1996. National Forest Policy of Malawi. Ministry of Natural Resources,Lilongwe, Malawi.

GoM, 1997. The Forestry Act 1997. The Malawi Gazette Supplement No. 5C.U.S.C.(1997). Government Printers, Zomba, Malawi.

GoM, 2001. Malawi National Forestry Program: Priorities for Improving Forestryand Livelihoods. Department of Forestry, Lilongwe, Malawi.

GoM, 2002. Department of Forestry Annual Report, 2001–2002. Ministry ofNatural Resources, Lilongwe, Malawi.

GoM, 2003a. Community Based Forest Management: A Supplement to the NationalForest Policy of Malawi, 1996. Ministry of Natural Resources and Environ-mental Affairs, Lilongwe, Malawi.

GoM, 2003b. Energy Policy for Malawi. Ministry of Energy and Mining, Lilongwe,Malawi.

GoM, 2007. Malawi Growth and Development Strategy: From Poverty to Prosper-ity 1996–2011. Government of Malawi, Lilongwe, Malawi.

Habitat, 1993. Application of Biomass-energy Technologies. United Nations Centrefor Human Settlements (Habitat), Nairobi.

Hiemstra-van der Horst, G., Hovorka, A.J., 2008. Reassessing the ‘‘energy ladder’’:household energy use in Maun, Botswana. Energy Policy 36 (9),3333–3344.

Hosier, R.H., Dowd, J., 1987. Household fuel choice in Zimbabwe: an empirical testof the energy ladder hypothesis. Resources and Energy 9, 347–361.

IFMSLP, 2007. Development of a Participatory Forest Management Plan forCustomary and State Forest Areas: A Guide for Frontline Staff. ImprovedForest Management for Sustainable Livelihoods Program (IFMSLP). IFMSLP,Lilongwe, Malawi.

Jere, P., Varela, K., Voysey, B., 2000. Synthesis study of initiatives in co-management of natural resources in Malawi. Working Group on Co-manage-ment of Forest Goods and Services. National Forestry Policy, Lilongwe, Malawi.

Kainja, S., 2000. Forest Outlook Study for Malawi. Department of Forestry,Lilongwe, Malawi.

Kalipeni, E., Zulu, L., 2002. From top-down to bottom-up: the difficult case of theBlantyre City Fuelwood Project. Journal of Southern African Studies 28 (1),117–136.

Kambewa, P., Mataya, B., Sichinga, W.K., Johnson, T., 2007. Charcoal – the reality:charcoal consumption, trade and production in Malawi. Technical Reportto the Forest Governance Learning Group. Malawi COMPASS II, Blantyre,Malawi.

Kayambazinthu, D., Locke, C., 2002. Forest management and diverse livelihoods inMalawi. LADDER Working Paper No. 24, University of East Anglia, Norwich./http://www.uea.ac.uk/dev/odg/ladderS.

Larson, A.M., Ribot, J.C., 2007. The poverty of forestry policy: double standards onan uneven playing field. Sustainability Science 2, 189–204.

ARTICLE IN PRESS

L.C. Zulu / Energy Policy 38 (2010) 3717–37303730

Leach, G., 1988. Residential energy in the Third World. Annual Review of Energy13, 47–65.

Leach, G., Mearns, R., 1988. Beyond the Woodfuel Crisis: People Land and Trees inAfricaEarthscan, London, England.

Lund, J.F., 2007. Is small beautiful? Village level taxation of natural resources inTanzania. Public Administration and Development 27, 307–318.

Lund, J.F., Treue, T., 2008. Are we getting there? Evidence of decentralized forestmanagement from the Tanzanian miombo woodlands. World Development 36(12), 2780–2800.

Mahiri, I., Howorth, C., 2001. Twenty years of resolving the irresolvable:approaches to the fuelwood problem in Kenya. Land Degradation & Develop-ment 12, 205–215.

Makungwa, S.D. 1997. Charcoal production study in the Blantyre Area. FRIM Report No.97003:ODC 867:891. Forestry Research Institute of Malawi, Zomba, Malawi.

MARGE, 2009. Malawi Biomass Energy Strategy, Consultancy study for Govern-ment of Malawi and European Union Partnership Dialogue Facility. Marcheageet Gestion de l’Environnement (MARGE), Lilongwe.

Masera, O.R., Saatkamp, B.D., Kammen, D.M., 2000. From linear fuel switching tomultiple cooking strategies: a critique and alternative to the energy laddermodel. Journal of World Development 28 (12), 2083–2103.

Michael-Phiri, M., 2006. Cashing in on the Charcoal Business. Nation Online, 19October 2006 /http:www.nationmalawi.com/print.asp?articleID=19133S.

Mugo, F., Ong, C., 2006. Lesson’s from Eastern Africa’s unsustainable charcoaltrade. Working Paper 20. ICRAF Working Paper no. 20. World AgroforestryCenter, Nairobi, Kenya.

Mwampamba, T.H., 2007. Has the woodfuel crisis returned? Urban charcoalconsumption in Tanzania and its implications to present and future forestavailability. Energy Policy 35 (8), 4221–4234.

NSO, 2005. Integrated Household Survey 2004–2005. National Statistical Office,Zomba, Malawi.

NSO, 2007. Monitoring the Situation of Children and Women. Malawi MultipleIndicator Cluster Survey 2007. Preliminary Report. National Statistical Office,Zomba, Malawi.

NSO, 2008. 2008 Population and Housing Census Preliminary Report. NationalStatistical Office, Zomba, Malawi.

Openshaw, K., 1997. Urban Biomass Fuels: Production, Transportation andTradingStudy. A consolidated report. Ministry of Energy and Mining, Lilongwe,Malawi.

Orr, B., Eiswerth, B., Finan, T., Malembo, L., 1998. Malawi Public Lands Utilization Study.University of Arizona and Forestry Research Institute of Malawi, Tuscon, Arizona.

Post, J., Snel, M., 2003. The impact of decentralised forest management on charcoalproduction practices in Eastern Senegal. Geoforum 34 (1), 85–98.

Ribot, J.C., Agrawal, A., 2006. Recentralizing while decentralizing: how nationalgovernments reappropriate forest resources. World Development 34 (11),1864–1886.

Ribot, J.C., 2009. Authority over forests: empowerment and subordination in Senegal’sdemocratic decentralization. Development and Change 40 (1), 105–129.

Satellitbild, 1993. Forest Resources Mapping and Biomass Assessment for Malawi.Sattelitbild/Department of Forestry, Lilongwe, Malawi.

SEI, 2002. Charcoal Potential in Southern Africa: CHAPOSA. Final Report. Stock-holm Environmental Institute, Stockholm, Sweden.

Shackleton, S., Campbell, B., Wollenberg, E., Edmunds, D., 2002. Devolution andCommunity-Based Management: Creating Space for Local People to Participateand Benefit?, Natural Resources Perspectives Overseas Development Institute,London.

World Bank, 1986. Malawi: Energy I Project (Second Wood Energy Project) StaffAppraisal Report. World Bank, Washington, DC.

World Bank, 1990a. The Malawi Charcoal Project Experience and Lessons. EnergySeries Paper no. 20. The World Bank, Washington, DC.

World Bank, 1990b. Malawi World Bank Economic Report on Environmental Policy– Aide Memoire, 17 August 1990. Unpublished Report.

Zulu, L.C., 2008. Community forest management in southern Malawi: solution orpart of the problem? Society and Natural Resources 21, 687–703

Zulu, L.C., 2009a. Politics of scale and community-based forest management insouthern Malawi. Geoforum 40, 686–699.

Zulu, L.C., 2009b. Land cover change, forest condition and community based forestmanagement in the Blantyre City Fuelwood Project (BCFP) area. In: Kalipeni, E.,Kakoma, I., Sanogo, Y.O., Fawcett, K., Warner, R.E. (Eds.), Turning Science intoAction: Biodiversity Conservation and Natural Resources Management inAfrica. Africa World Press, Trenton, NJ, pp. 251–298.