Printer Friendly

Understanding community criteria for assessing forest co-management programmes: evidence from Malawi/Comprendre les criteres communautaires pour evaluer les programmes de co-gestion: exemples concrets au Malawi/Comprension de los criterios comunitarios para la evaluacion de los programas de cogestion forestal: pruebas de Malawi.


Success and failure are terms commonly used in describing the status of forest co-management programmes (Axford et al. 2008, Pagdee et al. 2006). However, currently there are no globally agreed criteria for their measurement (Bowler et al. 2012, Pagdee et al. 2006). Therefore, the criteria necessarily vary depending on how success has been defined and understood and by whom (Pagdee et al. 2006).

A number of actors with diverse interests and goals are involved in the planning and implementation of forest co-management programmes, including community members (who are a diverse group in themselves), forest extension staff and donors. The different actors may perceive a forest co-management programme differently depending on their interest in the programme and socio-economic status (Castrol and Nielsen 2001). Therefore, depending on their perspectives, each actor group as well as individuals within the group, can identify different criteria for assessing co-management programmes and defend their chosen criteria as the correct one (Crook and Decker 2006, Pagdee et al. 2006, Pokharel and Suvedi 2007).

While acknowledging that all actors are important in forest co-management programmes, community members (both committee members and ordinary community members), are said to be the main actors at implementation level (Agarwal 2001, Pokharel and Suvedi 2007, O'Hara 2002) and are important in determining the programme's success or failure (Ostrom 1999). Therefore, identifying their perspective and opinions as to what constitutes successful co-management is crucial for reconciling conservation objectives and community development goals in the programmes (O'Hara 2002). Furthermore, identifying local perspectives facilitates the development of programme evaluation criteria that reflect elements that are important to local actors (who are usually the implementers and target population) as well as national or regional interests (Fraser et al. 2006). Additionally, limited understanding of local communities' perceptions, and opinions with regards to forest management approaches (including co-management), and lack of integration of local people's views in forest management plans could increase forest degradation (Weiss 2000). However, few studies have identified local criteria for assessing the success of forest co-management programmes and researchers often assess the effectiveness of programmes using criteria determined through literature reviews and based on the theoretical attributes of participatory management approaches in general, which at times do not reflect the interests of the participating actors (Pinkerton 1989, Pokharel and Suvedi 2007). Furthermore, forest co-management research and evaluations tend to be based on forest inventory data, with minimal consideration of the views of the local communities (Obiri et al. 2010). The few studies that have assessed local criteria have done so at the community level (e.g. Guthiga 2008, Pokharel and Suvedi 2007, Napier et al. 2005). Yet communities are heterogeneous in terms of values, social hierarchy and characteristics and as such individuals within a community may differ in their attitudes, opinions, understanding, definitions and views towards CFM (Larson et al. 2008, Mcfarlane and Boxall 2000, Mehta and Kellert 1998). Therefore, it is important to consider individual perceptions and understanding of co-management.

Using the case of a committee-led forest co-management programme of government forest reserves in Malawi, we present original quantitative and qualitative evidence on; 1) criteria for assessing the success of a forest co-management programme from local perspectives; 2) the differences and commonalities between ordinary community members and committee members in identifying criteria for assessing the success of a forest co-management programme and 3) the socio-economic factors influencing individuals' chosen criteria. This paper makes a novel contribution to the existing literature by improving our understanding of what local people consider to be important criteria for assessing forest co-management. Furthermore, this paper differs from previous studies, by going beyond community perspectives to further understand how different individuals within the community define criteria and what factors influence these differences. The knowledge gained will contribute to future design and evaluation of forest co-management programmes.


Study area

The study was conducted in Zomba Malosa (Zomba district) and Ntchisi (Ntchisi district) forest reserves. Zomba district is located in the southern region of Malawi and it covers a total of 2580 square kilometers. Agriculture (mainly maize) and forest resources form a large part of the economy and livelihoods for the majority of the population in the district. Zomba-Malosa forest reserve is the only gazetted forest in the district and covers an estimated area of 15,756 hectares, consisting of miombo woodlands and pine plantations. It is a catchment for major lakes and rivers in the country (e.g. Lake Malawi, Lake Chilwa, Lake Chiuta and the Shire River) and a significant source of water (both for domestic and agricultural use) in the district. The reserve is also a source of wood energy (charcoal and firewood) to households in the district as well as neighbouring districts. The forest reserve is located near a major road (M3 which connects to the country's central road-M1 at both ends, i.e. Balaka and Blantyre respectively), hence has easy access to markets for forest resources. This has accelerated deforestation and degradation of the reserve. Additionally, the reserve is being encroached upon in the peripheral areas for settlement and agriculture, resulting in further deforestation. It is estimated that 300 hectares of Zomba-Malosa forest reserve have been turned to agriculture land (Mauambeta et al. 2010).

Ntchisi district is located in the central region of Malawi and covers a total of 1 655 square kilometers. The agricultural sector is estimated to account for almost 80% of the district economy and livelihoods (Haarstad et al. 2009). The most common commercial crop grown in the area is tobacco, which requires a substantial amount of farmland, as well as wood resources for curing. Ntchisi district has 3 gazetted forest reserves, namely Ntchisi, Kaombe and Mndilasadzu forest reserves, with Ntchisi forest reserve being the largest covering an estimated 9,720 hectares. The reserve is located in a remote and rural part of the district, approximately 32 km from the district centre. The reserve is a source of non-timber forest products including mushrooms and edible caterpillars, and water for communities living around the reserve. Tree cutting in search of edible caterpillars is said to be a significant cause of deforestation and degradation in the reserve (personal communication, District forest extension officer 2012). Additionally, as a tobacco growing district, in Ntchisi tobacco farming is one of the major cause of deforestation, due to the high wood use for curing (e.g. Wiyo et al. 2015, Jumbe and Angelsen 2011).

Zomba Malosa and Ntchisi forest reserves are 2 of the 12 forest reserves where the Malawi Government is implementing the Improved Forest Management and Sustainable Livelihoods Programme (IFMSLP) through the Department of Forestry, with funding from the European Commission (Malawi Government 2007). The programme aims to address forest degradation and poverty through promoting community involvement in the management of State owned forest reserves. The programme is being implemented in phases; thus within the reserves, there are some blocks that are currently being co-managed by adjacent communities and the government after the signing of management agreements, and some blocks are still under state management as the process is still underway. To answer the questions in this study, the study sites had to fulfil the following criteria: 1) the forest block should be under full co-management, which means that communities living around the reserves have signed a management agreement with the government and are thus recognized as full participants, and 2) the programme should be sufficiently advanced, such that the participating communities have the potential to harvest and benefit from their designated forest block. Therefore, a total of 6 participating communities were surveyed in Zomba Malosa (Mtuluma, Jusu and Fikira) and Ntchisi (Nyanja, Nyanga and Mpamila) forest reserve.

Data collection

Both qualitative and quantitative methods were used for a more comprehensive inquiry and understanding of how different local actors define criteria for measuring success or failure of the co-management programme. The qualitative approaches included focus group discussions and key informant interviews, while individual interviews using a structured questionnaire were conducted to collect the quantitative data. A pilot survey to pre-test the questionnaire was undertaken in Lilongwe, in Dzalanyama forest reserves and with communities living around the forest reserve, before the start of the survey. Malawi Government, through the Department of Forestry, is implementing community forest management programmes with communities living around this reserve, but not under the IFMSLP. A total of 20 households were interviewed in the pilot survey. Therefore, pre-testing assisted in the revision of questions, standardizing the units of measurements and how questions were to be asked.

The researcher (LC) carried out the focus group discussions and key informant interviews and conducted the household surveys with the assistance of 3 field assistants who were trained on the survey questionnaire in the pilot survey location. Surveys were conducted in Chichewa, the first language of both respondents and interviewers, hence no interpreters were involved.

Key informant interviews and focus group discussions Meetings and interviews were held with 32 village leaders and committee members to gather general information for identifying the different local actor categories in the co-management programme at local level and their potential interests and objectives. A total of 21 focus group discussions were also done with community members in each study area, to obtain general qualitative information about the programme and possible success or failure criteria from the community. The discussions helped in identifying criteria or indicators representing a common underlying concept. Following Mendoza and Prabhu (2005) we identified the underlying concepts by firstly asking the groups to list all possible things or elements that they thought would indicate that the programme is a success and then naming all elements that they thought would indicate that the programme is a failure. Each element was further discussed in order to understand what each meant and to identify any linkages and connections, following which all responses and elements with a common meaning were put in one category. This was important because all identified criteria were used in regressions as dependent variables, and several items or criteria represent the same underlying concept are included in a regression as separate variables can lead to problems of multicollinearity and increased measurement errors (Hamilton 2004). Additionally, focus group discussions and key informant interviews further informed the development of the final individual questionnaire.

Questionnaire survey

The sampling frames for the household questionnaire surveys were based on village registers provided by the village heads. In cases where the list was unavailable from the village heads, a list of all households in the village was compiled by the researcher with the help of key informants. After verifying that the lists did not follow any particular order or social hierarchy (e.g. wealth status or kinship) every fourth household on the list was selected to form part of the study. A structured questionnaire was used to interview a total of 134 ordinary community members (74 male and 60 female) from 87 households were interviewed, representing approximately 33% of the households in the 6 sampled participating communities (see study area section). Ordinary community members comprise household heads and other adult members of the community who do not hold any leadership position in the programme. In each village, interviews were first conducted with household heads, who are usually male. Upon finishing the interview with the household head, a request was made to interview any other adult members of the household separately, on a different day. However, it was difficult to interview the spouse as many were not available on the agreed time and date; others simply declined to be interviewed. Hence, unequal numbers of household heads and other adults were interviewed.

In addition, we interviewed 21 committee members (15 male and 6 female). Committee members are local people in leadership positions and have greater formal contact with forest staff than ordinary community members through training and formal meetings (Malawi Government 2007). Furthermore, communities tend to select committee members who are elite in their society, e.g. wealthy individuals and the educated (Agarwal 2001, Vyamana 2009). Thus opinions and perceptions of committee members may differ from those of ordinary community members. Therefore, to capture any existing dynamics and differences in perceived criteria for measures of success of co-management programme because of actors' position or role in the programme at community level, the respondents to the face to face, were therefore grouped into committee members, and ordinary community members.

The questionnaire included both closed and open ended questions. However, open ended questions were used to elicit information on respondents' perceived criteria for assessing success or failure of co-management (see supplementary material). The responses were later coded according to the criteria categories identified during the focus group discussion. The questionnaires also gathered information on respondents' basic socio-economic and demographic characteristics including age, gender, major income sources, education level.

Data analysis

Chi-square tests were used to test if perceived criteria for measuring success or failure of co-management differed between ordinary community members and committee members. To explore socio-economic factors that determine individuals' perceived criteria for measuring success or failure of the co-management programme, logistic regression models were used (1). The logistic regression model presents the log-likelihood of the explanatory variables on the success criteria, and was used with an assumption that each of the criteria was mentioned independently of the other, i.e. the choice of one criterion does not influence the choice of another. According to Wooldridge, (2002) the logit regression equation is specified as:

Logit (Y = 1) = [[beta].sub.0] + [[beta].sub.1district] + [[beta].sub.2wealth] + [[beta].sub.3gender] + [[beta].sub.4education] + [[beta].sub.5landsize] + [[beta].sub.6foreslivelihood]

The dependent variables were drawn from the responses on what criteria respondents would use to measure success or failure of the co-management programme. Dummy variables (1=yes, 0=no) were created for each of the criteria identified by respondents. For some criteria (i.e. conserved forest and improved livelihoods), a number of indicators with the same underlying principle were combined to form a category, so as to avoid multicollinearity and increased measurement errors. The internal consistency of the categories was further measured by the reliability coefficient, Cronbach alpha (Cronbach 1951), which ranges from 0 to 1. The larger the value, the greater the reliability of the scale (Cronbach 1951), the Cronbach alpha values for conserved forest and improved livelihoods were 0.7637 and 0.6254, respectively suggesting that responses combined to represent these criteria are reliable.

A number of studies have found or hypothesised that socio-economic factors including age, gender of respondent, major income sources, education level and community (2) affect community members' attitudes towards co-management (e.g. Htun et al. 2011, Jalilova et al. 2012). Therefore, the explanatory variables in our logistic regression are: district or community, wealth indicator, gender of respondent, number of years in school, land size and forest based livelihood source. Tests for multi-collinearity were done for the different explanatory variables included in the models. The Variance Inflation factor (VIF) scores of [less than or equal to] 2 and Tolerance ranging from 0.67 to 0.81 indicate a weak correlation between the explanatory variables (Allison 1999). Bootstrapping (1000 resamples) was used in estimating the coefficients (Wooldridge 2002). All the data analyses were conducted using STATA version 11.2.


Perceived criteria for measuring success (or failure) of co-management

A total of five criteria for assessing the success of the co-management programme were identified by ordinary community members, in both districts. These were; conserved forest, access to forest resources, community participation in decision making and management, establishment of community development infrastructure and improved livelihoods (Table 1). Criteria were either represented by a single response or by multiple responses. Participation in decision making and management; better access to forest resources (fuel wood and NTFP), were each represented as a single response in both districts. Similarly, common to both sites, the criterion 'development projects and infrastructure' was represented by a single element, namely construction of roads. However, conserved forest and improved livelihoods as criteria for measuring success were represented by multiple responses with a common underlying concept. Conserved forest as a criterion was represented by the following responses; no or reduced deforestation, increase in number of trees and regrowth/seedlings and saplings and reduced tree felling. Improved livelihoods as a criterion was represented by the following responses; increase in income level, increase in livelihood sources, increase in employment opportunities, establishment of income generating activities, provision of credit services; food security and improved knowledge and skills in both forest management and entrepreneurship.

Perceived criteria for measuring success of the programme were significantly different between ordinary community members and committee members in both Zomba Malosa and Ntchisi districts. Ordinary community members' emphasis was on access to forest resources and improved livelihoods, whilst committee members' emphasis was on forest conservation (Table 1). Similarly, when asked which one was the most important indicator, the majority of ordinary community members highlighted either access to forest resources or improved livelihoods, whilst the majority of committee members highlighted forest conservation, and the difference was statistically significant in both Zomba Malosa ([chi square] = 11.79, p = 0.036) and Ntchisi districts ([chi square] = 8.97, p = 0.042) (Table 2).

A majority of ordinary community members in Zomba Malosa (68%) and Ntchisi (63%) indicated that their perceived criteria are based on household goals, whereas, a majority of committee members in Zomba Malosa (76%) and Ntchisi (62%) indicated that their perceived criteria are based on programme goals (Figure 1), and these differences were significant ([chi square] = 13.51 p = 0.001).

Further to observed differences in perceived criteria between ordinary community members and committee members, comparison of means between ordinary community members and committee members shows that, in Zomba Malosa committee members are more educated, older, wealthier (p = 0.1), and with larger land holding sizes than ordinary community members (Table 3). Similarly, in Ntchisi district, comparison of means shows that committee members are significantly older (p = 0.01), wealthier (p = 0.1), educated (p = 0.01) and with larger land holding sizes (p = 0.01) than ordinary community members (Table 3).

Determinants of perceived indicators for measuring success of the programme

The likelihood of perceiving conserved forest as a criterion is significantly lower for respondents in Ntchisi than those in Zomba Malosa (p = 0.05) and significantly increased with increasing wealth status (p = 0.10), number of years in formal education (p = 0.01) and the size of land owned (p = 0.001) (Table 4). The odds of perceiving access to forest and forest resources as a criterion is higher for respondents in Ntchisi (p = 0.001) and female respondents (p = 0.01) and decreases significantly with increasing land holding size (Table 4). Perceiving community participation in decision making and management activities as a criterion is significantly higher for male respondents (p = 0.001), respondents with more years in formal education (p = 0.01). There is no evidence to suggest that any socio-economic characteristics significantly influence the perception of community development projects as a criterion for measuring success of the programme (Table 4). However, perceiving improved livelihoods as a criterion significantly decreases with increasing wealth indicator scores (p = 0.001) and number of years in formal education (p = 0.1).


Whilst ordinary community members in Zomba Malosa and Ntchisi identified five criteria for measuring success of forest co-management programmes, committee members identified only three (Zomba Malosa) and four (Ntchisi). Conserved forest and improved livelihoods were the two criteria common to both ordinary members and committee members. These are also the overall goals and objectives of the forest co-management programme (Malawi Government, 2007). However, the two groups differed in the priority accorded to these two criteria, with committee members more likely to prioritise forest conservation. In addition, community participation in decision making; access to and availability of forest resources; and infrastructure development, are also important criteria for assessing a co-management programme from an ordinary community member's perspective. However, these are usually not included in forest management impact assessment studies.

Perceived criteria differed significantly between ordinary community members and committee members, which is consistent with the findings of Pokharel and Suvedi (2007). In theory anyone in the participating community can be elected as committee member (Chinangwa et al. 2015); however, Table 3 shows that committees in both Zomba Malosa and Ntchisi are dominated by the elite members of community. Thus suggesting that the difference in perceived criteria for measuring success of forest co-management programme, between ordinary community members and committee members, may be due to the pre-existing differences in their socioeconomic status rather than their position in the programme. Additionally, the observed difference in perceived criteria and the basis for defining the criteria between ordinary community members and committee members, suggest that the committee members are not representative of, responsive to and downwardly accountable to, their constituents. This is further evidenced by community members reporting a lack of formal or informal community forest management meetings with committee members. Furthermore, we were unable to access any records of meetings in either district. Similar findings were reported by Chinangwa et al. 2015, who observed that although committee members are expected to conduct meetings in their community, none of the committee members in their study area reported to have carried out any meetings. This indicates a lack of fora where the views of ordinary community members could be gathered by the committee members, and also where committee members could be held accountable by their constituents. Additionally, a study in the same area by Chinangwa et al. (2015), found that committee members' accountability and representativeness of their constitutes is limited because of unclear rules with regards to election procedure (e.g. lack of clearly defined terms of office and schedules for the next election). Furthermore, some committee members were not elected but were just appointed by either government officials or village leaders (Chinangwa et al. 2015), hence further reducing accountability and responsiveness of the committee as such members are loyal to the interest of those who appointed them and not of the community or their constituents (Ribot 2004 and 2003, Oyono 2003). The failure to represent ordinary community members may therefore limit the programme's ability to address their interests and goals effectively, thus alienating them from the programme and potentially resulting in conflicts (Negendra 2007, Zulu 2013).

Interestingly, criteria identified by ordinary community members were considerably closer to the specific programme objectives. (3) This could suggest that ordinary community members were just mimicking what was presented to them by government during the introduction and sensitization stages of the programme. However, the initial activities (i.e. sensitization meetings), were done with individuals, especially those in leadership positions, with the intention that they will in turn sensitize their constituents, yet the majority of community members indicated that they had never attended any forest co-management meetings. Therefore, it is plausible that the criteria are considerably closer to the programme objectives, because the programme designs took into account the needs of local people either based on literature or experiences from earlier projects in Malawi. Thus the community forestry management programme in Malawi may be evolving from the first generation community forestry which focused on structural issues such as tenure, protection, and regulation, to second generation that gives attention to issues such as equity, benefit sharing, and the wider livelihoods impacts (e.g. Lawrence 2007). However, the realisation of these objectives (i.e. equity, benefit sharing and livelihoods) may be undercut by the programme's institutional choices as it fails to empower the ordinary community members to demand and access these benefits.

Community participation is one of the principal components of a co-management programme, and as such is an important factor in the programme's success or failure (e.g. McDermott and Schreckenberg 2009). Interestingly, none of the committee members identified community participation as a criterion. Similar findings have been shown by Pokharel and Suvedi (2007), who also found that ordinary members are more likely to support participation in governance and decision making of forest programmes as criteria for success than are programme leaders. Committee members are already in decision making positions, therefore less likely to consider participation in decision making and management as a criterion of success, as they may therefore view this criterion as already having been achieved. Additionally, committee members may believe that their powers and privileges (e.g. training) will be reduced if ordinary members actively participate in decision making, hence less likely to mention it as a criterion of success (e.g. Zulu 2010 and 2008). Ordinary community members are not in decision-making positions and usually not included in decision making meetings (e.g. Chinangwa et al. 2015), therefore would appreciate the ability to participate and influence decisions in the programme, to ensure that their interests, such as access to a programme's benefits and accountability of the leaders, are heard and/or addressed (McDermott and Schreckenberg 2009, Mmehta and Heinen 2001, Gillingham and Lee 1999).

Ordinary community members perceived access to forest and forest resources, and improved livelihoods as major criteria for measuring the success of the programme. This is expected because forest resources form an essential part of rural livelihoods; therefore, attaining legal access to forest resources is an important benefit and one of the major reasons for communities' participation in co-management programmes (Cronckleton et al. 2012, Jumbe and Angelsen, 2006). However, in mentioning access to forest and forest resources as a criterion, communities did not imply unrestricted harvesting of forest products: the focus group discussions revealed that communities in both districts would prefer that access to forest should be controlled so as to allow continued use of the resources by the current and future generations. This is also evident as some ordinary community members identified conserved forest as a criterion for measuring success of the co-management programme. Nevertheless, there were some differences within and across the discussion groups with regards to who should control access, as some members indicated that they would prefer if control of access were returned to government forest staff, whilst others indicated that the committee members should maintain control. Groups that had women only and the youth was leaning toward returning it to the governments, whilst the mixed groups (where largely the men speak), were more into retaining the control under committee. This reflects inequalities in benefit sharing and resource utilization rights across the different gender groups within the communities even under co-management. Thus supporting findings by Mawaya and Kalindekafe (2007), who noted that in Malawi women's empowerment to exercise the utilization rights is limited, even under a participatory forest management programme.

Community development projects and infrastructure (specifically, the development of a road network) was also highlighted as an important criterion of success of the co-management across both study sites, but by a minority of respondents. A road network, like any other community development project, benefits the whole community. Therefore, it is plausible that only a few ordinary community members identified it as criterion, as a majority of them stated that they defined the criteria based on individual or household goals, not village goals.

Logistic regression shows that respondents in Ntchisi are less likely to perceive conserved forest as a criterion than those in Zomba Malosa. Poteete and Ostrom (2004) suggest that among other factors, community members may be more willing to participate in conservation or motivated to conserve forests if they perceive their forest as degraded. If they perceive the forest resources as abundant, they may see no reason for restricting usage or employing strict conservation measures. Ntchisi forest reserve has more tree counts per plot (mean = 27 mature tree/plot), than Zomba Malosa forest reserve (mean = 11 mature trees/plot) (Chinangwa 2014) and Zomba Malosa forest reserve is relatively more degraded than Ntchisi forest reserves. Therefore, for community members in Zomba Malosa, conservation of the available trees and improvement of the forest condition is a higher priority and thus a more important criterion for measuring success of the forest co-management programme. Consequently, a community's perception of their forest's condition could have an effect on the outcomes of the co-management programmes, as communities who perceive their forest to be degraded may respond to the management and utilization rights given to them by taking charge to conserve the forest, whilst those that perceive their forest resources as abundant may take advantage of the utilization right by harvesting, which if not done sustainably could result in forest degradation.

Individuals with larger land holdings are more likely to perceive forest conservation and less likely to perceive access to forest and forest resources as criteria. Individuals with large land holdings can ably meet their livelihood needs from agriculture and may own more private trees on their land than those with small land holdings; hence less dependent on resources from the forest reserve and likely to identify with the conservation objectives of co-management (Reij and Waters-Bayer 2001). However, for individuals with small land holdings, forest resources are an important livelihood source supplementary to agriculture (Poteete and Ostrom 2004), thus they are more likely to favour the programme if it allows for access and utilization forest resources.

Female respondents are more likely to identify access to forest and forest resources as a criterion for measuring success of forest co-management than male respondents. The difference is due to differences in resource use and extraction among different gender groups (Colfer and Capistrano 2005). In the study area, women are largely responsible for collection of firewood and NTFP, and more involved in forest based livelihood strategies than men. Female respondents are less likely to perceive participation in decision making and management as a criterion. This is perhaps expected, as due to cultural norms and practices, female members of the community rarely assume decision making positions and rarely contribute during public decision making forums (e.g. Mawaya and Kalindekafe 2007). This was also noted in the mixed gender focus group discussion, as women rarely contributed, unless specifically requested by the facilitator (hence we also held women-only group discussions). Therefore, cultural norms and practices influence local perceptions and attitudes towards forest management programmes including co-management approaches (Shackleton et al. 2002). However, it does not imply that women do not want to contribute to decision making, but it highlights the need for establishment of appropriate forums that could enable the women and other marginalized groups in society to be heard and benefit from the programme effectively.

The logit results show that individuals with more years of formal education are more likely to identify forest conservation as a criterion. This supports the claim that formal education enhances positive perceptions towards forest conservation in an individual (e.g. Samdin et al. 2010). However, knowledge of forest conservation and its benefits could also be enhanced through other informal training, awareness meetings, contact with conservation experts (e.g. forest extension staff) and transfer of local knowledge among community members both in time and space (Charnley et al. 2008, Bhattarai et al. 2005). Therefore, informal training and meetings with regards to forest and forest management should be encouraged in order to enhance conservation objectives of the programme. Finally, individuals with more years formal education are less likely to perceive improved livelihoods as a criterion for assessing the success of the forest co-management programme. This is attributed to the fact that attaining a formal education is positively associated with individuals' ability to get employment and diversify their income sources (Hatlebakk 2012), hence the educated are less likely to depend on the forest as their major source of livelihood and income. The wealth indicator also shows a similar trend, that richer individuals are more likely to identify forest conservation as a criterion, but less likely to identify improved livelihoods as criteria for measuring success of co-management. This further helps explain why committee members perceive conserved forest as a criterion for measuring the success of forest co-management programme, because committee members tend to be the elite in their society, e.g. wealthy individuals and the educated (Agarwal 2001, Vyamana 2009), thus are more focused on conservation aspects of the programme than the livelihood aspect.


Five criteria for measuring the success of the programme were identified by respondents: improved livelihoods, access to forest resources, community participation in decision making and management, conserved forest and establishment of community infrastructure. A majority of ordinary community members measure the success of the co-management programme based on how the programme is addressing or can meet their individual goals. From a local perspective, a successful forest co-management programme should address both the utilitarian value of the forest and ecological conservation. In addition to livelihoods and forest conditions indicators, assessments of forest co-management should thus include access to forest resources, community participation in decision making and the establishment of community infrastructure, which are reflective of local actors' goals and success indicators. However, the programme's unclear institutional processes with regards to the election of committee members and accountability measures undercut the programmes progress in effectively realizing these goals and outputs. Indicators for measuring the success of forest co-management differ with individual characteristics and actor group. Committee members' priorities differed from those of ordinary community members as a whole, but were more reflective of wealthier and better-educated members of the community. This highlights a potential lack of accountability in co-management programmes and the potential for community institutions to be captured by local elites. The lack of accountability could potentially compromise the success of co-management programmes.

A community's perception of the status of the forest may affect their criteria and the outcomes of the programme, as communities may interpret and respond to the rights that are transferred to them according to their perceptions of forest condition. Community perceptions of criteria for measuring success of programmes may be determined by a number of household socio-economic characteristics, including district/ community, gender of household heads, wealth status and education. Evaluation and impact assessment studies should be designed to capture perspectives and experiences across social strata (e.g. gender, wealth) within a community. Cultural norms and practices influence local perceptions and attitudes towards forest management programmes. Therefore, co-management programmes should be able to create appropriate fora that will enable the marginalized in society to be heard and benefit from the programme. Furthermore, public hearing and public audit sessions in the co-management programme should be introduced to enhance committee member's accountability and responsiveness to their communities. Furthermore, for a programme to be effective it must be understood and implemented within the existing local social, cultural, economic and ecological status or environment. Finally, the study suggests that education enhances individual's attitudes toward conservation outcomes, hence implementation training and capacity building that integrates traditional and local knowledge with regards to forest conservation and utilization are recommended.


This work was supported by the International Centre for Research in Agroforestry (ICRAF). We acknowledge the staff of the Department of Forestry, Malawi, Mr T. Kamoto, Mrs S. Gama, Mr. A. Munyenyembe, Mr Magagula, Mr Nangwale, Mr Makupete, Mr Goneta, and all ground staff, for their assistance during the field work. We also acknowledge the support of our research assistants, Mr W. Chinangwa, Mr. M. Munyenyembe and Mr. G. Chipofya. Finally, we acknowledge the households and communities in our study sites, Zomba Malosa and Ntchisi who gave generously of their time.


AGARWAL, B. 2001. Participatory exclusions, community forestry, and gender: An analysis for south Asia and a conceptual framework. World Development 29(10): 1623-1648.

ALLISON, P.D. 1999. Logistic regression using the SAS system theory and application. Sage, SAS Institute, North Carolina.

AXFORD, J.C., HOCKINGS, M.T. and CARTER, R.W. 2008. What constitutes success in pacific island community conserved areas? Ecology and Society 13(2): 45-60.

BHATTARAI, B., OJHA, H. and HUMAGAIN, Y. 2005. Is leasehold forestry really a pro- poor innovation in Nepal? Evidences from Karve District. Forest and Livelihoods 4(2): 17-30.

BOWLER, D., BUYUNG-ALI, L., HEALEY, J.R., JONES, J.P.G., KNIGHT, T. and PULLIN, A.S. 2012. Does community forest management provide global environmental benefits and improving local welfare? Frontier Ecology and Environment 10(1): 29-36.

CASTROL, A.P. and NIELSEN, E. 2001. Indigenous People and co-management: Implications for Conflict Management. Environmental Science and Policy 4(4): 229-239.

CHARNLEY, S., FISCHER, A.P. and JONES, E.T. 2008. Integrating traditional and local ecological knowledge into forest biodiversity conservation in the Pacific Northwest. Forest Ecology and Management 246(1): 14-28.

CHINANGWA, L. 2014. Do co-management programmes reconcile community interests and forest conservation: a case study of Malawi. Phd thesis submitted to Bangor University.

CHINANGWA, L., SINCLAIR, F., PULLIN, A.S. and HOCKLEY, N. 2015. Can co-management of government forest reserves achieve devolution? Evidence from Malawi. Forest, Trees and Livelihoods. 80/14728028.2015.1087886

COLFER, C.J.P. and CAPISTRANO, D (eds.). 2005. The politics of decentralization: forests, power and people. Centre for International Forestry Research, earthscan, London. 1 pp.

COULIBALY-LINGANI, P., SAVADOGO, P., TIGABU, M. and ODEN, P.C. 2011. Decentralization and community forest management in Burkina Faso: Constraints and challenges. International Forestry Review 13(4): 476-486.

CRONBACH, L.J. 1951. "Coefficient alpha and the internal structure of tests". Psychometrika 22(3): 297-334.

CRONKLETON, P., PULHIN, J.M. and SAIGAL, S. 2012. Co-management in community forestry: How the partial devolution of management rights creates challenges for forest communities. Conservation and Society 10(2): 91-102.

CROOK, B. and DECKER, E. 2006. Factors affecting community-based natural resource use programs in Southern Africa. Sustainable Forestry 22(3-4): 111-133.

FRASER, E.D.G., DOUGILL, A.J., MABEE, W.E., REED, M. and MCALPINE, P. 2006. Bottom up and top down: analysis of participatory processes for sustainability indicator identification as a pathway to community empowerment and sustainable environmental management. Environmental Management 78(2): 114-127.

GILLINGHAM, S. and LEE, P.C. 1999. The impact of wildlife-related benefits on the conservation attitudes of local people around the Selous Game Reserve, Tanzania. Environmental Conservation 26(3): 218-228.

GUTHIGA, P.M. 2008. Understanding local communities' perceptions of existing forest management regimes of a Kenyan Rainforest. Social forestry 1(2): 145-166.

HAARSTAD, J., JUMBE, B.L., CHINANGWA, S., MPONELA, P., DALFELT, A. and DISCH, A. 2009. Environmental and socio-economic baseline study-Malawi. NORAD. 235910?-.

HAMILTON, L.C. 2004. Statistics with Stata. Version 8, Pacific Grove, Duxbury Press, Belmont, California.

HATLEBAKK, M. 2012. Regional variation in livelihood strategies in Malawi. South African Journal of Economics 80(1): 62-76.

HTUN, N.Z., NOBUYA-MIZOUE, N. and YOSHIDA, S. 2011. Determinants of local people's perceptions and attitudes toward a protected area and its management: a case study from Popa Mountain Park, Central Myanmar. Society & Natural Resources 25(8): 1-16.

JALILOVA, G., KHADKA, C. and VACIK, H. 2012. Developing criteria and indicators for evaluating sustainable forest management: A case study in Kyrgyzstan. Forestry Policy and Economics 21: 32-43.

JUMBE, C.B.L, ANGELSEN, A. 2011. Modeling choice of fuelwood source among rural households in Malawi: a multinomial probit analysis. Energy Economics 33: 732-738

JUMBE, C.B.L, ANGELSEN, A. 2007. Forest dependence and participation in CPR management: Empirical evidence from forest co-management in Malawi. Ecological Economics 62: 661-672.

LARSON, A.M. 2005. Natural resources and decentralization in Nicaragua: Are local governments up to the job? World Development 30(1): 17-31.

LAWRENCE, A. 2007. Beyond the second generation: Towards adaptiveness in participatory forest management. CAB Reviews: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural Resource 2(28): 1-15.

MALAWI GOVERNMENT. 2007. Baseline report: Improved Forest Management for Sustainable Livelihoods (IFMSLP). Department of Forestry, Lilongwe.

MAUAMBETA, D.D.C., CHITEDZE, D., MUMBA, R. and GAMA, S. 2010. Status of forests and tree management in Malawi. CURE MALAWI. en/2015-02.../strategy-documents?

MAWAYA, C. and KALINDEKAFE, M.P. 2007. Access, control and use of natural resources, in southern Malawi: A gender analysis. In FLINTAN F. and TEDLA, S. (eds) Natural Resource Management: The Impact of Gender and Social Issues. Fountain Publisher, Ottawa. 88 pp.

MCDERMOTT, M.H. and SCHRECKENBERG, K. 2009. Equity in Community Forestry: Insights from North and South. International Forestry Review 11(2): 157-170.

MCFARLANE, B.L. and BOXALL, P.C. 2000. Factors influencing forest values and attitudes of two stakeholder groups: The case of the Foothills Model Forest, Alberta, Canada. Society and Natural Resources 13(7): 649-661.

MEHTA, J.N. and KELLERT, S.R. 1998. Local attitudes towards community-based conservation policy and programmers in Nepal: A case study of the Makalu-Barun conservation area. Environmental Conservation 25(4): 320-333.

MENDOZA, G.A. and PRABHU, R. 2005. Combining participatory modelling and multi-criteria analysis for community-based forest management. Forest Ecology and Management 207: 145-156.

MMEHTA, J.N. and HEINEN, J.T. 2001. Does community-based conservation shape favourable attitudes among locals? An empirical study from Nepal. Environment and Management 28(2): 165-177.

NAPIER, v., branch, G. and HARRIS, J. 2005. Evaluating conditions for successful co-management of subsistence fisheries in Kwazulu-Natal, South Africa. Environmental Conservation 32: 165-177.

NEGENDRA, H. 2007. Drivers of reforestation in human-dominated forests. Proceedings of the National Academy of Science 104(39): 15218-15223.

O'HARA, P. 2002. Community forestry liberation through scaling down our failures. Forests, Trees and People Newsletter 46: 46-53.

OBIRI, J.A.F., HALL, J.B. and HEALEY, J.R. 2010. Degradation, species invasion and management responsibility in Tanzanian dry forests: What do local people say? In BONGERS, F. and TENNIGKEIT, T. (eds.) Degraded Forests in Eastern Africa. Earthscan, London. 305 pp.

OSTROM, E. 1999. Coping with tragedies of the commons. Annual Reviews, Political Science 2: 493-535.

Oyono, P.R. 2003. Institutional deficit, representation and decentralized forest management in Cameroon. Elements of natural resource sociology for social theory and public policy. Working Paper No.15, Environmental Governance in Africa Series. The World Resource Institute, Washington, DC.

PAGDEE, A., KIM, Y.S. and DAUGHERTY, P.J. 2006. What makes community forest management successful: A meta-study from community forests throughout the world. Society & Natural Resources 19(1): 33-52.

PINKERTON, E. 1989. Co-operative management of local fisheries: new directions for improved management and community development. Vancouver, Canada: UBC Press.

POKHAREL, R.K. and SUVEDI, M. 2007. Indicators for measuring the success of Nepal's community forestry program: Local perspective. Human ecology 14(1): 68-75.

POTEETE, A.R. and E. OSTROM. 2004. Heterogeneity, group size and collective action: The role of institutions in forest management. Development and Change 35(3): 435-461.

REIJ, C. and WATERS-BAYER, A. 2001. Farmer innovation in Africa: a source of inspiration for agricultural development. Earthscan, London.

RIBOT J. 2004. Waiting for Democracy. The politics of choice in Natural resources decentralization. WRI report. Washington, D.C.

RIBOT, J.C. 2003. Democratic decentralization of natural resources: institutional choice and discretionary power transfers in Sub-Saharan Africa. Public Administration and Development 23(1): 53-65.

SAMDIN, Z., YUHANIS, A.A., RADAM, A. and YACOB, M.Y. 2010. Factors influencing the willingness to pay for entrance permit: The evidence from Taman Negara National Park. Journal of Sustainable Development 3(3): 212-220.

SHACKLETON, S., CAMPBELL, B., WOLLENBERG, E. and EDMUNDS, D. 2002. Devolution and community-based management: Creating space for local people to participate and benefit? natural resources perspectives. Overseas Development Institute, London.

VYAMANA, V. 2009. Participatory forest management in the Eastern Arc Mountains of Tanzania: Who benefits? International Forestry Review 11(2): 239-253.

WEISS, T. 2000. 'Beyond peasant deforestation: Environment and development in rural Jamaica'. Global Environmental Change: Human and Policy Dimensions 10(4): 299-305.

WIYO, A.W., FIWA, W. and MWASE, W. 2015. Solving Deforestation, Protecting and Managing key water catchments in Malawi using smart public and private partnerships. Journal of Sustainable Development 8: 251-261.

WOOLDRIDGE, J. 2002. Econometric analysis of cross sectional and panel data. MIT Press, London.

ZULU, L.C. 2010. The forbidden fuel: Charcoal, urban wood fuel demand and supply dynamics, community forest management and wood fuel policy in Malawi. Energy policy 38: 3717-3730.

ZULU, L.C. 2008. Community forest management in Southern Malawi: Solution or part of the Problem? Society and Natural resources 21(8): 687-703.

ZULU, L.C. 2013. Bringing people back into protected forests in developing countries: insights from co-management in Malawi. Sustainability 5: 1917-1943.

L. CHINANGWA (a,b&c), A.S. PULLIN (d) and N. HOCKLEY (c)

(a) Ministry of Agriculture, Department of Land Resources Conservation, P.O. Box 30291, Lilongwe 3, Malawi

(b) United Nations University, Institute of Advanced Study of Sustainability, 53-70 Jingumae 5-Chome, Shibuya-ku, Tokyo 150-8925, Japan

(c) School of Environment, Natural Resources and Geography, Bangor University, Bangor, Gwynedd, LL57 2DG, UK

(d) Centre for Evidence-Based Conservation, School of Environment, Natural Resources and Geography, Bangor University, Bangor, Gwynedd, LL57 2DG, UK


(1) Only data from ordinary community members were included in the regression models.

(2) In this study community is represented by district (i.e. Zomba or Ntchisi).

(3) The four specific objectives of the programme: (1) promotion of sustainable livelihood strategies within impact areas; (2) enhancing equitable access to forest resources by increasing the area under sustainable forest management arrangements; (3) strengthening governance of key forest resources; and (4) enhancing communication and advocacy among stakeholder groups within the forest sector (Malawi Government 2007).

Caption: FIGURE 1 Percentage response ofcommittee members and ordinary community members with respect to whatforms the basis of their perceived criteria for measuring
TABLE 1 Criteria for measuring success of forest co-management
programme, as perceived by ordinary community members
and committee members in Zomba and Ntchisi district

                                    Percentage response (%)

Criteria                         Zomba                  Ntchisi

                         Ordinary   Committee   Ordinary   Committee
                         members     member     members     member

Conserved forest           25.4     96.7 ***      36.5     89.1 ***
Access to forest           67.7       0 **        86.3      9.8 **
  (Fuel wood and NTFP)
Participation in           7.3          0         4.1          0
  decision making
  and management
Development projects       4.9         6.6        4.3         5.8
  and infrastructure
Improved livelihood        58.9      33.7 *       66.8      35.3 *

Note: *** = 1% significant difference; ** = 5% significant difference;
* = 10% significant difference

TABLE 2 Most important criterion for measuring success of forest
co-management programme, as perceived by community members in Zomba
and Ntchisi district

                                 Percentage response (%)

Criteria                      Zomba                  Ntchisi

                      Ordinary   Committee   Ordinary   Committee
                      members     member     members     member

Conserved forest        15.2       66.7        6.5        59.1
Access to forest        41.7         0         56.3        9.8
  resources (Fuel
  wood and NTFP)
Participation in        7.3          0         4.1          0
  decision making
  and management
Development             4.9         6.6        4.3         5.8
  projects and
Improved livelihood     30.9       26.7        28.8       25.3

TABLE 3 Comparison of means of individual characteristics variables
between ordinary members and committee members


Variable                     Ordinary   Committee   t-statistic
                             members     members

Age of household head         41.78       42.39        -0.39
  (in years)
Land size (in hectors)         0.7        0.88         -0.41
Household size (number of      4.89       4.78          0.44
  adults and children)
Number of years in school      4.83       5.55         -1.53 *
Wealth indicator               7.34       7.44         -0.45
  (ordinal scale, 4-11)


Variable                     Ordinary   Committee   t-statistic
                             members     members

Age of household head         39.81       43.12      -1.74 **
  (in years)
Land size (in hectors)         0.65       1.43       -1.64 ***
Household size (number of      5.52       5.21        1.09
  adults and children)
Number of years in school      4.64       6.78       -3.57 ***
Wealth indicator               6.87       7.75       -1.74 *
  (ordinal scale, 4-11)

a = Significance levels * = 10%; ** = 5%; *** = 1%

b = Wilcoxon-Mann Whitney test was used for the non-parametric
variables (wealth indicator).

TABLE 4 Factors influencing individual perceived criteria for
measuring success or failure of co-management programme
(coefficients from logistic regression)

                         Conserved   Access to    Participate in
                          forest       forest     decision making
                                     and forest   and management

District (1=Ntchisi,     -0.34 **    1.31 ****         0.57
0=Zomba)                  (0.25)       (0.27)         (0.38)

Wealth indicator          0.07 *       -0.04           0.04
(ordinal scale, 4-11)     (0.03)       (0.03)         (0.05)

Gender (1=female,          0.08       0.96 ***      -1.47 ****
0=male)                   (0.30)       (0.29)         (0.42)

Age (in years)             0.01        -0.01           -0.02
                          (0.01)       (0.01)         (0.02)

Number of years          0.10 ***       0.01         0.14 ***
in school                 (0.03)       (0.03)         (0.05)

Land size                0.09 ****    -0.05 *          0.01
(in hectares)             (0.03)       (0.03)         (0.03)

Forest based               -0.16        0.05           0.13
livelihoods               (0.36)       (0.34)         (0.42)
(1=yes; 0= no)

Constant                 -1.84 ***      0.88        -3.61 ****
                          (0.61)       (0.63)         (0.94)

Prob > [chi.sup.2]         0.03         0.00           0.00

Number of                   134         134             134

Log likelihood            -229.22     -218.10         -140.43

Pseudo [R.sup.2]           0.05         0.13           0.13

                          Development     Improved food
                          projects and      security

District (1=Ntchisi,          0.05            -1.58
0=Zomba)                     (0.74)          (0.32)

Wealth indicator             -0.16          -0.14 ***
(ordinal scale, 4-11)        (0.12)          (0.04)

Gender (1=female,            -0.25            -0.07
0=male)                      (1.01)          (0.32)

Age (in years)                0.04            -0.01
                             (0.03)          (0.01)

Number of years              -0.13           -0.07 *
in school                    (0.11)          (0.04)

Land size                    -0.08            -0.04
(in hectares)                (0.09)          (0.04)

Forest based                 -0.52            0.23
livelihoods                  (1.13)          (0.38)
(1=yes; 0= no)

Constant                     -3.06            -0.59
                             (1.91)          (0.66)

Prob > [chi.sup.2]            0.08            0.00

Number of                     134              134

Log likelihood               -40.31          -180.74

Pseudo [R.sup.2]              0.14           0.1188

a. * = Significance levels (* = 10%; ** = 5%; *** = 1%; **** = 0.01%)

b. All coefficients and standard errors are boot strapped
COPYRIGHT 2017 Commonwealth Forestry Association
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2017 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Chinangwa, L.; Pullin, A.S.; Hockley, N.
Publication:International Forestry Review
Article Type:Report
Date:Mar 1, 2017
Previous Article:Optimism, hopes and fears: local perceptions of REDD+ in Nepalese community forests/Optinisme, espoirs et craintes: perceptions locales de la REDD+...
Next Article:Looking beyond tenure in China's Collective Forest Tenure Reform: insights from Yunnan Province, Southwest China/Regards au-dela du bail dans la...

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters