hope and disappointment : The dynamics of expectations in conservation and development pilot projects

We explore the dynamics of expectations in international forest conservation and development programs, and the impacts and implications of (unfulfilled) expectations for actors involved. Early stages of new international conservation and development programs, often involving pilot projects designed to test intervention concepts at village level, are characterized by large amounts of resources and attention, along with high expectations of success. However, evidence shows that these early expectations are rarely fulfilled. Despite this repeated pattern and growing engagement with expectations in critical conservation and development literature, little is known about the dynamics of expectations in conservation and development pilot projects. We address this knowledge gap first by exploring concepts from the sociology of expectations. We then unpack expectations in a case study of REDD+ pilot projects in Tanzania, using extensive qualitative data reflecting the perspectives and experiences of a wide range of actors involved. Our study finds that expectations play a performative role, mobilizing actors and resources, despite uncertainty identified among policy-makers and practitioners. We also find that once raised, expectations are dynamic and continually mediated by actors and social contexts, which conflicts with attempts to ‘manage’ them. We argue therefore that a trade-off exists between fully piloting new initiatives and raising expectations. We also argue that failure to address this trade-off has implications beyond pilot project objectives and timelines, which are experienced most acutely by village communities. We argue for more critical engagement with expectations and the embedding of accountability for expectations in conservation and development practice. Our findings also challenge the discourse of ‘needing’ to pilot, which prioritizes awareness, impact and innovation without fully considering the potential negative impact of unfulfilled expectations. 2018 The Authors. Published by Elsevier Ltd. This is an openaccess article under the CCBY license (http:// creativecommons.org/licenses/by/4.0/).


Introduction
Large scale, internationally-led programs have repeatedly been framed as necessary solutions to forest governance challenges in the tropical Global South (Adger, Benjaminsen, Brown, & Svarstad, 2001;Mace, 2014;Redford, Padoch, & Sunderland, 2013). Programs have included integrated conservation and development programs (ICDPs), participatory forest management (PFM), payments for ecosystem service (PES) and, most recently, reducing emissions from deforestation and forest degradation, and enhancing forest stocks through improved forest conservation and man-agement (REDD+). These programs are designed to tackle environmental problems such as biodiversity loss and climate change, as well as to support community development challenges. These problems and challenges are defined in such a way that technical, multiple-win and increasingly market-based solutions are required to solve them (Igoe & Brockington, 2007;Li, 2007). The early stages of these new programs, often involving village level pilot projects, are characterized by large amounts of money, resources, attention and high expectations (Fletcher, Dressler, Büscher, & Anderson, 2016;Redford et al., 2013). The reality of these initiatives rarely lives up to the high early-stage expectations, and so subsequent solutions that require new policy models and technical programs are sought (Li, 2007;Mosse, 2005;Redford et al., 2013). As such, a number of academics have conceptualized these programs as 'conservation fads' (Fletcher et al., 2016;Lund, Sungusia, Mabele, & Scheba, 2016), defined by Redford et al. (2013, p. 437) as 'approaches that are embraced enthusiastically and then abandoned'.
Discussion about the relationship between these international programs and expectations is increasing amongst both practitioners and academics, with expectations defined as imagined ideas about the future that circulate through social interaction (Konrad, 2006;Van Lente, 2012). Most recently, the focus of this discussion has been on REDD+, which has stalled at the end of the pilot phase leaving many early expectations unfulfilled (Lund et al., 2016;Sills et al., 2014). It is argued that the early stages of REDD+ have led to the development of an 'economy of expectations', whereby local level processes, realities and visions for the future are altered through involvement with global, marketbased conservation programs (Dressler, 2017;Fletcher et al., 2016). As a result, expectations and the management of expectations have been highlighted as one of the biggest challenges of REDD+ pilot project implementation (Atela, 2015;Sunderlin et al., 2014), with practitioners now required to 'develop strategies to deal with the backlash' (Fletcher et al., 2016, p. 674).
Despite this interest in expectations, there has been little detailed exploration of how they are produced, how they circulate and the impact they have throughout the different stages of conservation and development projects. Thus more understanding is needed about what the science and technology (STS) literature calls the 'sociology', or 'dynamics', of expectations in this context (Brown & Michael, 2003;Van Lente, 1993). Exploring the dynamics of expectations is necessary for a full understanding of social change (Borup, Brown, Konrad, & Van Lente, 2006). We posit that this is particularly relevant in relation to pilot projects, which are often used to test new international conservation and development programs at the local level in order to generate quick and tangible results, to influence policy and to generate further donor funds (Adams, 2003;Garí, 2013;Vreugdenhil, Slinger, Thissen, & Rault, 2010). As such, pilot projects require buy-in, engagement and action from all of the actors involved and so drive social change. Yet they rarely come with a guarantee of continued funding and activity post-pilot.
We address this knowledge gap first by reviewing sociology of expectations literature, which is primarily drawn from the field of STS. We identify core themes and characteristics that are relevant to international forest conservation and development programs. We then use this to investigate a case study of REDD+ pilot projects in Tanzania, drawing primarily on in-depth narrative interviews conducted with actors from global to village levels and collected after the pilot projects were phased out. This includes detailed investigation of two pilot project case studies, which provide empirical evidence of contrasting approaches to pilot projects and expectations. Through this analysis, we contribute to a better understanding of the dynamics of expectations in conservation and development practice, as well as the dynamics of expectations more broadly. In doing so, we also contribute to a deeper understanding of pilot projects, and other interventions, as agents and outcomes of social change. This approach contrasts with common, instrumental forms of project evaluations that focus on the impacts of interventions on forest governance or performance against project objectives (Li, 2007;Mosse, 2005). As such we also provide new and useful insights to policy-makers and practitioners involved in international forest conservation and development projects.

The sociology of expectations
Expectations can be defined as imagined ideas about the future that are produced, circulated and mediated through social interaction, resulting in social change (Berkhout, 2006;Konrad, 2006;Van Lente, 2012). Actors' actions and decisions are always made in relation to expected outcomes and consequences (Berkhout, 2006;Van Lente, 2012). Expectations can be both positive and negative, and both individual and collective, and are therefore both contextspecific and related to broader shared or collective visions (Konrad, 2006). Collective expectations develop in relation to shared 'imaginaries', which are defined as 'imagined forms of social life and social order that centre on the development or fulfilment of innovative scientific and/or technical projects' (Jasanoff et al., 2007). New conservation and development programs are often framed within a multiple-win rhetoric, towards imaginaries of international forest governance for the benefit of all (Igoe & Brockington, 2007). And it is argued that market-based solutions such as REDD+ heighten these dynamics due to their emphasis on future speculation and their transnational, abstract nature, which is less aligned with local contexts than previous programs (Dressler, 2017).
Early stages of new innovation or technological development both drive and are driven by hyper expectations, or hype, which can be defined as unreasonable and unachievable expectations of what the new innovation can deliver (Brown, 2003). As such, a sense of urgency ensues, driven by both fear of environmental harm and the imaginaries of future conservation (Brown & Michael, 2003;Büscher & Dressler, 2007). Newness is fetishized and as such ideas that are framed as being new, different and distinct are favored over the advancement of existing solutions, not least because shortfalls associated with past solutions are erased (Brown & Michael, 2003;Brown, 2003). As such, Mosse (2004, p. 640) argues that in development 'the intense focus on the future, on new beginnings, is rarely moderated by an analysis of the past'. These new beginnings often require the use of show or pilot projects to bring new policy to life (Igoe & Brockington, 2007). Hype and expectations can therefore focus energy and attention on one new solution, becoming a barrier to critical thinking, to alternative solutions and to approaches that favor incremental change (Brown, 2003).
Expectations can be described as being performative in that their existence mobilizes both actors and resources, and as such they provide an important function in the early stages of innovation (Borup et al., 2006;Brown & Michael, 2003;Konrad, 2006). They can coordinate and broker relationships between a wide range of actors -both horizontally (for example between policymakers) and vertically across different scales from the global to the local (Borup et al., 2006). As collective expectations develop, 'communities of promise' build up around them (Brown, 2003, p. 5) and actors join these discursive communities despite individual uncertainties and reservations, often to ensure that they do not get left behind (Konrad, 2006;Van Lente, 2012). In this sense, economies of expectation can develop in which new realities are created (Borup et al., 2006). Expectations have thus been defined as 'forceful presence' (Van Lente, 2012, p. 773). This forceful presence can be seen in the context of global, market-based conservation mechanisms that create new social structures, nature valuations and imaginaries that in turn encourage more activity and higher expectations (Dressler, 2017;West, 2006).
There is much discussion in both the STS literature and critical conservation and development literature about the level of intention of raising expectations in relation to their performative role. On the one hand, it can be framed as an inevitable and unavoidable outcome of social interaction and innovation (Konrad, 2006). Once something is enacted, it becomes part of a reality that is both linked to the actor's original intentions but also combines with other actors and contexts to take on a life of its own that often results in unintended consequences (West, 2006). However, others argue that innovators and policy-makers deliberately raise expectations in order to mobilize resources and enrol actors into com-munities of promise (Brown & Michael, 2003;Sung & Hopkins, 2006), particularly in conservation and development where actors such as NGOs and government agencies have to compete for scarce resources, such as donor funds, legitimacy and reputation (Li, 1999). Elevated expectations, created by hype associated with early stages of innovation, results in hype and disappointment cycles (Borup et al., 2006). Actors' efforts to sustain expectations are overwhelmed by the reality of underlying issues and so communities of promise collapse (Brown, 2003;Van Lente, 1993). This results in what Mosse (2004) refers to as an unintended but inevitable gap between international development policy and the realities of implementation. Repeated cycles of new international conservation and development programs, or 'fads' (Redford et al., 2013) therefore result in repeated hype and disappointment cycles. Disappointment can then lead to outcomes including apportioning blame, disillusionment, damaged credibility of innovators and policy-makers, and adverse effects on future innovations (Brown & Michael, 2003;Brown, 2003;Sung & Hopkins, 2006;Van Lente, 1993). Such outcomes could include conservation and development NGOs losing their legitimacy (Dressler, 2017), resistance to future projects at the local level (Leach & Scoones, 2015;Li, 2007) and environmental destruction by villagers whose expectations of project involvement have not been met (West, 2006). However, in some cases new cycles of hype provide a protected space for new innovation and past disappointment is forgotten (Borup et al., 2006;Konrad, 2006). Although cycles of expectation and disappointment can be conceptualized as inevitable, their impacts and implications are highly contextual, related to the social dynamics of expectations.
The initial framing of the innovation by those developing and/or selling it impacts the development of collective and individual expectations (Sung & Hopkins, 2006). However, expectations are continually mediated by actors' past experiences, social interactions, networks and activities, and social framings (Brown & Michael, 2003;Leach & Scoones, 2015;Sung & Hopkins, 2006). For example, West (2006) finds that villagers engage with projects with the understanding that they are entering into long-term, reciprocal, social relationships with practitioners towards imaginaries of development and progress. This results in disappointment once projects end and these expectations are not met. Expectations, uncertainty and disappointment can be conceptualized as dynamic, continually influencing and being influenced by social discourse and interactions (Konrad, 2006). As such, attempts by practitioners to manage expectations once they have been raised are likely to be unsuccessful (Weszkalnys, 2008).
A relationship between actors' proximity to the production of knowledge, and their levels of uncertainty and expectations, can also be identified. Brown and Michael (2003) find that actors closest to the production of knowledge (such as innovators and policymakers) have high levels of uncertainty about the success of the new idea or solution and, as a result, low expectations. Actors furthest away from knowledge production (such as medical patients and recipients of development projects) tend to have low uncertainty and therefore the highest levels of expectations. Those closest to knowledge are the source of raised expectations, yet disappointment affects the user groups furthest away from knowledge, highlighting the asymmetrical nature of negative impacts of unrealistic expectations (Brown & Michael, 2003;Van Lente, 2012). In Tanzania, evidence suggests that the REDD+ pilot phase fell well short of initial expectations and promises of change (Benjaminsen, 2014;Lund et al., 2016;Svarstad & Benjaminsen, 2017). As such, Brown (2003) argues that a reworking of economies of expectation is required in order that the uncertainties of those closet to knowledge become more transparent; particularly to those most negatively impacted by hype and disappointment cycles.

The case study of REDD+ pilot projects in Tanzania
REDD+ pilot projects in Tanzania are used as an instrumental case study or a bounded case that is explored in detail in order to illustrate an issue of concern (Creswell, 2012). REDD+ is a mechanism developed by the United Nations Framework Convention on Climate Change (UNFCCC) based on the principle that through carbon markets, or international donor funding, developing countries are financially rewarded for preventing deforestation, protecting forests and hence increasing global carbon stocks. When the initial pilot stages commenced in many countries following the Bali Action Plan in 2007, there were high hopes internationally that REDD+ would achieve multiple wins by contributing to global climate change mitigation targets, biodiversity protection and local forest conservation and development objectives (Phelps, Friess, & Webb, 2012;Visseren-Hamakers, McDermott, Vijge, & Cashore, 2012).
Large amounts of funding and resources were employed to get countries 'REDD+ ready' and pilot projects implemented in order to test REDD+ mechanisms (Lund et al., 2016;Sunderlin et al., 2014). Critical voices also emerged during the early stages, with actors including academics, indigenous groups and practitioners warning of potential human rights, land tenure and justice issues (Chhatre et al., 2012;Clements, 2010;Corbera, 2012;Lang, 2010) As the REDD+ ready phase and associated pilot projects continued, it became evident that the mechanism was harder to implement than expected and that global REDD+ funding mechanisms were not yet in place (Lund et al., 2016). As such, many projects have stalled, been abandoned, or have evolved into more traditional conservation and development projects that no longer focus on monetary incentives for carbon storage and sequestration . REDD+ therefore makes a timely and relevant case study through which to explore expectations in the early stages of forest conservation and development initiatives.
In Tanzania, the REDD+ (known locally as MKUHUMI) readiness phase was active between 2009 and 2014. It was supported by US $80 million of bilateral funding from Norway's International Climate and Forest Initiative (NICFI) and managed by the Norwegian Embassy in Tanzania (Kaijage & Kafumu, 2016). Additional national-level strategic support was given by the UN and World Bank Forest Carbon Partnership Facility (REDD-Desk, 2017). The Norwegian funding supported a national REDD+ Task Force made up primarily of government actors, thematic working groups consisting of government, civil society and private sector actors, and a REDD+ Secretariat based at the University of Dar es Salaam. These institutions supported the Vice President's office (VPO) in the development of the Tanzanian REDD+ strategy document (VPO, 2013). Funding was also used to support a large number of research projects and to enable the implementation of nine pilot projects, seven of which reached completion. A mix of international NGOs and well-established national NGOs were chosen to implement the projects, which lasted between four and five years. The objectives of the pilot projects included testing REDD+ mechanisms with communities in a wide range of contexts, getting communities ready for REDD+, delivering widespread stakeholder awareness and involvement in REDD+, delivering REDD+ results such as emission reduction, and supporting national policymaking (NIRAS, 2015).
The NGOs took very different approaches to piloting REDD+, with some aiming to meet all objectives and fully test the mechanism and others choosing to focus on only a few elements. Two individual case studies were chosen to explore expectations in REDD+ pilot projects in more detail and to reflect two contrasting approaches to piloting. Tables 1 and 2 present the key facts related to these two individual pilot project case studies. It is argued that the pilot projects achieved some objectives, generated useful insights and demonstrated that REDD+ at the village scale in Tanzania is feasible (NIRAS, 2015). However aside from investment into the National Carbon Monitoring Centre (NCMC), continued funding via donor support or carbon finance was not in place when the pilot projects were completed and national plans for a phase two of REDD+ were not clear (Lund et al., 2016). As such, most of the pilot projects ended with little scope for continuation through REDD+, although many of the NGOs involved continued working with the communities through other projects and funding.

Research design
An interpretive, actor-orientated approach to case study research was taken (Long, 2003). Methodologically this involves using ethnographic methods to unpack lived experiences from the perspectives of individual actors and actor groups, and emphasizes the interplay between outside influences such as internationally-led interventions, and the different realities, perceptions, social interests, and relationships of actors involved (ibid.). We agree with Cooper and Pratten (2015) that ethnography should draw heavily on individual narrative in order to do justice to participants' lived experiences. This article is based primarily on 70 in-depth, narrative interviews conducted with a wide range of different actors involved in the pilot projects. These narratives were collected between September 2015 and May 2016 and were selected to reflect a broad range of respondent demographics, characteristics and viewpoints and included international, national, regional, district and village-level actors. Additional ethnographic data was collected during this period, as well as during additional visits to Tanzania between September and October 2014 and March and August 2015, in order to support the narrative interviews and ensure credibility of the research (Lincoln & Guba, 1985). The two individual pilot project case studies outlined in Tables 1 and 2 were chosen to provide rich data or 'thick' description (Geertz, 1994) of project level and village actor and experiences, and in order to reflect two very different approaches to piloting. Data was collected in two villages involved in each pilot project. They are referred to as K1 and K2 in Kilosa and R1 and R2 in Rungwe for confidentiality reasons. These villagers were selected following key informant interviews with national and district NGO representatives, regional and district government representatives and academics, and consulting NGO project documents. They were selected as villages that had been most fully involved in the pilot projects. Table 3 summarizes the data collected.
Most of the data, including the narratives, was collected after the pilot projects had ended, with the objective of gathering reflections of the whole pilot process, as well providing insights into the impact of pilot projects beyond their completion date. It is noted that the expectation narratives represent the actor framings at that Table 1 Kilosa project key facts. 1

Implemented by Tanzania Forest Conservation Group (TFCG) and Mtandao wa Jamii wa Usimamizi wa Misitu Tanzania (MJUMITA) -Tanzania Community Forest Network
Piloted with villages that had had few previous (non-governmental) interventions Community-managed forests and community-based forest management (CBFM) Active for five years (until December 2014) Aimed to demonstrate a pro-poor approach to improved village forest management through international carbon financing Fully tested REDD+ mechanisms and aimed to get communities 'REDD ready' Followed global carbon standards and processes with the aim to complete a project document and gain verification for future carbon trading Free Prior Informed Consent (FPIC) consultation undertaken at the start of the project Tested benefit-sharing mechanisms by completing a trial carbon payment (two trial payments were planned but only one completed due to technical issues) Established Community Based Forest Management (CBFM), which involved village committees, land use plans and establishment of Village Land Forest Reserves (VLFRs) for protection under REDD+ Livelihood projects including conservation agriculture and micro-finance groups Carbon measurement, reporting and verification (MRV) High awareness of MKUHUMI and wider project among villagers. High levels of participation Main messages communicated to villagers: potential negative impacts of deforestation and degradation, information about carbon and REDD+ including potential benefits At time of data collection new sustainable charcoal project started in Village K2 with plans to expand to K1. Local NGO staff supporting the communities to continue with livelihood projects where possible, but no additional donor support for REDD+ projects 1 Compiled from own data and NIRAS (2015).  (Brown & Michael, 2003). It was also noted that at the start of data collection, the lead researcher and research assistants were automatically linked to REDD+ by villagers, particularly in Kilosa. We tried to overcome this by spending time prior to each interview explaining our position as independent academic researchers, however we note that a belief that we may have been able to directly influence REDD+ may have influenced some of the responses. The data was analyzed in two phases: inductively then deductively. Firstly, narrative content analysis was used in order to understand the experiences of actors and the meanings they attributed to these experiences (Elliott, 2005). Storylines around expectations were identified within each narrative, with storylines defined as part of a narrative that allows actors to 'give meaning to specific physical or social phenomena' (Hajer, 1995, p. 56). The actor storylines were then compared and analyzed inductively to find patterns. During the second phase, these narratives were analyzed using the aforementioned dynamics of expectations concepts.

The early stages of REDD+ in Tanzania: hype, urgency and expectations
When reflecting back on the early stages of the REDD+ pilot projects in Tanzania, national and international actors identified high levels of hype or hyper expectations, which are highlighted by Brown (2003) as being a core characteristic of early innovation. International actors saw REDD+ as an opportunity for Tanzania to establish its position internationally as a leader in REDD+ knowledge and practice. Expectations of continuation post-pilot were identified among national actors, including government officials and NGOs. These expectations of continuation included more funding from donors, a national level REDD+ program spearheaded by the government, and continued funding for communities via carbon markets, and were largely related to the 'opportunity for communities to benefit, to take carbon as one of the products of the forest'. 1 There were also expectations that REDD+ would provide a solution to forest conservation in Tanzania and become a source of much-needed, ongoing, financial support for the forestry department. One government official reflected that 'people thought 'ah okay, the forests are now safe because of REDD"... even my director had that notion.' 2 REDD+ was framed as a multiple-win solution to forest governance issues, producing collective expectations of a market-led solution towards imaginaries of abundant resources and well-protected forests (Igoe & Brockington, 2007;Leach & Scoones, 2015;Lund et al., 2016). This was despite the fact that funding was secured for a pilot phase only, demonstrating the over-inflated expectations inherent in the hype during the early stages of new innovation (Brown, 2003).
A sense of urgency (Brown & Michael, 2003;Büscher & Dressler, 2007) was also identified by national actors when reflecting on the early stages of the projects, which was cited by Embassy employees as influencing the decision to have NGOs lead the pilot projects (as opposed to the Tanzanian government). One of the actors involved in the development of the pilot projects reflected that 'the whole international thing' 3 driving REDD+ meant that the pilot projects began before key actors such as NGOs and government officials were fully aware of what REDD+ involved. Part of the original donor strategy was to build on the existing PFM tradition in Tanzania but despite this, much of the emphasis was put on the new elements of REDD+, particularly in relation to carbon payments. The NGOs 'were encouraged to include front-loaded payments within their budgets to test payment and benefit sharing arrangements in the expectation of making longer-term carbon sales' (Blomley et al., 2016, p. 2). This aligns with Brown and Michael (2003)'s argument that by emphasising newness, innovation gains more traction, funds and attention, and that this in addition drives higher expectations. This also allowed the REDD+ project to be seen as a new beginning and so failures of past programs could be overlooked (Igoe & Brockington, 2007;Mosse, 2004).
This early-stage hype and associated expectation influenced and informed the activity of many of the national actors during the early stages of the REDD+ pilot projects. This included development of a national media campaign to raise awareness of REDD+, and the framing of the pilot projects by the NGOs: this forest will pay you [the villagers]. Not only for one year, but will pay you continuously! I mean this is a bank account. You are saving money and getting interest. So I think [there was] an expectation that OK by the time the project is ending, we will have our project document, we will have our process verified, and we've qualified to be paid. 4 In Kilosa, a range of actors including villagers, local leaders, district government and local NGO staff reflected on high expectations at the start of the process, during which they were visited by the Task Force, the Embassy and the NGO, and were involved in the FPIC process. These expectations were both in relation to outcomes of the pilot project itself, such as village education and development, and assumptions of future benefits of REDD+, such as ongoing carbon payments and improved local climate conditions. During the early stages, negative expectations related to the project were also identified among villagers in both K1 and K2. These included worries that 'these Europeans have come from their home countries to come and steal our land', 5 that people would be moved from their farms and that wild animals would be introduced to the area. As such, strong positive and negative expectations existed alongside one another during the initial stages of the Kilosa pilot projects (Konrad, 2006). Factors such as negative past experiences of government conservation programs and proximity to strict national wildlife parks drove these negative expectations, or fears, illustrating the mediation of expectations by actors and contexts (Konrad, 2006;Sung & Hopkins, 2006). Narratives of actors involved in the REDD+ pilot project in Rungwe reflect very different expectation dynamics at the start of the project. At this time expectations in relation to carbon payments began to develop among regional, district and village government actors as a result of national media campaigns, attendance of meetings on REDD+ and visits from the Task Force. However, the implementing NGO decided not to focus on REDD+ specific mechanisms such as trial carbon payments, preferring to focus on research and livelihood activities. This was largely as a result of concerns about 'making promises to communities that you can't deliver' 6 and maintaining their legitimacy among communities with whom they have a longstanding relationship. The NGO did not go through a village-wide consultation process at the start of the project, gaining consent from village leaders, and did not focus on widespread participation in the livelihood projects. Broader village participation was encouraged in the education programs, which framed the project and forest conservation as being about reducing the risk of drought, floods and rising temperatures, 7 without directly focusing on carbon. As a result, awareness of and participation in the REDD+ pilot project among actors outside of village government and committees was low, and village level actors spoke of very few expectations of the pilot project. Expectations can therefore be conceptualised as a product of the framing of those who are 'selling' the new idea (Sung & Hopkins, 2006); in this case the NGOs in Kilosa and Rungwe, who took very different approaches to the pilot projects.

The performative function of expectations
The performative function of expectations (Borup et al., 2006;Brown & Michael, 2003;Konrad, 2006;Sung & Hopkins, 2006) can be identified through national level and Kilosa actor narratives, driving and being driven by the aforementioned international and national level hype, sense of urgency, and assumptions of a REDD+ future. Actors were enrolled into 'communities of promise' (Brown, 2003, p. 5), both horizontally at the national level, and also vertically across regional, district and village levels in Kilosa. At the international and national level, these discursive communities of promise developed despite personal uncertainties of a number of actors. These uncertainties were largely related to the unknown nature of carbon financing mechanisms central to REDD+ . Differences between lower personal levels of expectations and much higher collective expectations at the time were identified (Konrad, 2006). Reflections of a number of national level actors suggest that collective expectations were performative in that they enrolled actors into communities of promise and project engagement despite their personal concerns. There were also suggestions that they resulted in an uncritical approach to piloting, as identified by Brown (2003) 'I sometimes feel guilty that I was part of it. You know, some people [were] just preaching like priests, the way they preach about God and Jesus Christ and all those kind of things, but without having a critical analysis about what it really means.' 8 At the international and national level uncertainty was in fact performative, acting as one of the driving forces behind the choice to pilot, in order to avoid 'policy-making in a void '. 9 The framing of REDD+ as new, unknown and filled with future possibilities mobilized a large amount of funding and activity and led to the aforementioned perceived need to pilot to bring this new policy to life (Igoe & Brockington, 2007).
In Kilosa, expectations were also performative and can be seen to be both the cause and outcome of change within the district and the villages (Borup et al., 2006;Van Lente, 2012). Communities of promise built up around the expectations of village development, ongoing carbon payments, and improvements to local ecosystem services (through better forest conservation). Actors within these communities of promise, who included the Task Force, district officials and local leaders, reassured villagers that they would not be moved from their farms, that wild animals would not be brought into the area, and that the 'future is bright and REDD'. 10 Expectations were then in turn influenced by the early stages of project activity, which included villagers receiving their first trial payment, the building of the office and the establishment of some of the livelihood activities: 'What made me change [from objecting] is that I received education that they will not keep animals again, we will just conserve forest and water sources only... also another thing is after seeing that they were supporting the construction of this office and also they promised us they will sell the carbon dioxide and we can get money that will help to conserve our forest and do village developments.' 11 Although village-level actors had been made aware of the project timescales during the FPIC process, longer term expectations related to village development and carbon payments began to rise. This in turn led to expectations becoming what Van Lente (2012, p. 773) calls 'forceful presence'. An'economy of expectation' developed, with new social structures, discourses and activity emerging, further driving collective expectations of considerable change (Dressler, 2017). As part of the CBFM land planning process, Village Land Forest Reserves (VLRFs) were established, the size of which was decided by village committees. However, due to the requirements and promises of REDD+, the communities were encouraged to 'take on larger areas of forest under reservation than they would otherwise have done '. 12 Once the VLFR had been gazetted, committees and leaders in both villages asked people with farms in the reserve areas to leave (for more on this see Vatn, Kajembe, Mosi, Nantongo, & Silayo, 2017). This illustrates the performative role of expectations in the displacement of villagers that has been highlighted in relation to other conservation and development programs (Büscher & Dressler, 2007;Fairhead, Leach, & Scoones, 2012). In K1, where the relocations affected more people than in K2, this has resulted in conflict between village leaders and people refusing to move from their farms in the VLFR. This conflict was a central part of many actor narratives in K1, with villagers split between those who support the moves and those who feel it was unfair. This conflict was continuing at the time of data collection (village leaders estimated about 25 farmers continued to farm in the VLFR), with threats of violence reported by both parties and farmers being taken to court. 13  Kilosa. This could be explained by the fact that they did not include any of the displaced families in their sample. Data collection timings may also be a factor. They collected data in 2013 while the project was still underway and so the disappointments associated with the project ending, and the continued impact of the economy of expectations in the villages -including conflicts over farms -are likely to have impacted the way in which people frame their experiences. changed for villagers as a result of the REDD+ pilot projects in Rungwe.

Were expectations raised intentionally?
Among policy-makers, project implementers and other national-level actors, different framings of intentionality in relation to expectations can be identified, which aligns with different framings in the STS literature. Some non-NGO actors, who were not directly involved with implementation of pilot projects, reflect Brown and Michael (2003) and Sung and Hopkins (2006) in suggesting that expectations were raised intentionally in order to change behavior among local actors, with one national actor claiming that: 'Some [NGOs] took the whole concept of carbon credit... as a way to encourage communities to engage in forest management, and to me that was a false promise'. 14 Conversely, other actors, including the implementing NGO in Kilosa, framed expectations as being an unintended but unavoidable consequence of piloting REDD+. For example, NGO practitioners reflected that it was hard to communicate the complex concept of REDD+ to communities in a way that ensured their full understanding but didn't raise expectations. This aligns with the view of Konrad (2006) that expectations are an inevitable product of social interactions and processes. FPIC is one such process, which is intended to deliver full disclosure of all aspects of the project including benefits, challenges and information about carbon and the carbon markets, in order that communities are equipped and empowered to accept or reject the project (Kibuga et al., 2011). The final report commissioned by the donor claimed that FPIC 'generated many advantages, among which managing expectations and mitigating future risks were the most important' (NIRAS, 2015, p. 20). It is argued that the fact that some villages in Kilosa rejected the project demonstrates the effectiveness of FPIC (Vatn et al., 2017). However, a number of actors reflected that in reality despite its good intentions, the FPIC process actually increased expectations among villagers: 'We studied one criteria called [free] prior informed consent. One is to be willing without being influenced based on what he sees in the village. But actually they were influenced by being told that REDD will bring you this money.' 15 Some NGO practitioners raised concerns with FPIC, challenging its ability to be effective in communicating the complexity of REDD +, echoing broader discussions about the limitations of the instrument (Mahanty & McDermott, 2013). These reflections on FPIC, along with other project elements implemented in good faith such as trial carbon payments, expose an inevitable link or trade-off inherent in piloting. This is a trade-off between fully piloting new initiatives, which involves securing high levels of awareness, engagement and participation, and raising expectations. The comparison between the Rungwe project where the NGO did not achieve high levels of awareness and engagement but experienced few of the negative impacts of expectations, and the Kilosa project in which the NGO achieved high levels of awareness, engagement and expectations, emphasizes the need for recognition of this trade-off and its potential consequences for villagers.
This trade-off can be positioned as a product of the broader dynamics of conservation and development. Actors such as NGOs are required to compete for scarce resources, which requires them to sell future success to donors and recipients alike (Dressler, 2017;Li, 1999;Mosse, 2005), or as one NGO representative put it 'this is the way the system works -we always write overoptimistic proposals because you [the donors] demand it from us!' 16 Innovative projects that showcase the new mechanism fully and achieve high levels of awareness and involvement are judged to be a success (Büscher, 2014;Igoe & Brockington, 2007), with the final REDD+ pilot project evaluation reports judging the Kilosa project to have been much more of a success than the Rungwe project (see NIRAS, 2015). Raising expectations among villagers may not have been intentional, but it is nonetheless an inevitable consequence of fully piloting new programs, particularly in relation to market-based mechanisms that are built around speculative future benefits (Dressler, 2017).

Hype and disappointment
The actor narratives at the national level reflect a general pattern of rising expectations that then fell significantly over time as the reality of issues and challenges became clear. Lack of political will among government officials, lack of donor support post-pilot and low carbon prices were identified by national and international actors as the main causes of the decline in expectations. This pattern follows the hype and disappointment cycles identified in the STS literature (Brown & Michael, 2003;Brown, 2003;Konrad, 2006). A number of national actors spoke about their disappointment that REDD+ had not lived up to its high expectations. Most of the disappointment however was expressed in relation to villagers. When data was collected between six and 18 months after the end of the pilot projects, only a few national actors spoke about continued expectations of REDD+, and none spoke about experiencing ongoing negative personal impacts. National and international actors had moved on to other projects and programs, and many commented that they had not engaged with REDD+ for some time.
At the time of data collection in Kilosa, the pilot projects had been completed over a year previously and there were no plans for continuation of the REDD+ mechanism (e.g. carbon payments). The villagers received only one trial payment of the expected two, largely as a result of issues with measurement, and the project had not got to a stage where it could be verified. One NGO employee explained that in regards to REDD+ 'Kilosa's luck has faded out.' 17 Despite this, village level actor narratives did not wholly reflect a hype and disappointment cycle and different narratives could be identified. Firstly, some actors did not feel any disappointment due to their perception that the project had bought many benefits to them and the village as a whole. These actors were predominately those who had been heavily involved in the project, whether as village leaders, committee members or livelihood project participants. To them the project had 'woken up' 18 the villagers and had brought much-needed development and education, and improved forest conditions. A second group of actors, comprised of those less involved in the project and those affected by the farm relocations in K1, expressed a strong sense of disappointment in the project: 'I personally don't feel good... before I thought well of them, that maybe our village is going to benefit, but for now I see this MKU-HUMI issue hasn't any benefit to me '. 19 This disappointment was largely in relation to the lack of continued carbon payments, a feeling of injustice that the project only benefitted a few people, and the continuing conflict over farm relocations. For some villagers these land issues were framed as a core part of the legacy of the project, especially when reflecting that 'if 14 20 The contrasting ways in which the project was framed within actor narratives reflects their contrasting experience, but may also be influenced by the way in which they perceived themselves in relation to the project, expectations and the researcher. By framing the project as highly successful, those who benefitted most from it were able to legitimise their roles within the economy of expectations, shoring up their position in relation to future projects. Similarly, the narratives of those who felt they had not benefitted from the project reflect their experiences, including their struggles within the economy of expectations and the desire to benefit more in the future. These narratives were told in light of ongoing expectations with regards future carbon payments, which had continued despite the pilot project ending: '...they [the villagers] haven't given up, but you find that when we go to the public meetings they normally discuss that we were told that we'd be paid every year. [They ask] what's going on? Therefore we normally answer them that after it is being measured it's taken to the world market there and they [MKUHUMI] have their process of discussing it so that the money can be paid... it takes time...' (Interviewer) 'So are you still expecting the second payment?' 'Yes, that is our hope, because that is what they had promised us' 21 West (2006, p. 197) states that 'people make claims when something is at stake'. Through the process of narrative interviewing, village actors may have been making claims over any future carbon payments, thus positioning themselves in relation to the economy of expectations.
The negative impacts associated with hype and disappointment cycles can therefore be seen to be asymmetrical (Brown & Michael, 2003), as one national NGO representative reflected: 'For NGOs it's annoying when you lose money and maybe have to lay off some staff but they're professional and they'll go off and get another job somewhere else. This is the way the world works. But those communities that we went out to and said 'hey this is a new opportunity -and now we can't make it happen for you.' That I think is really bad.' 22 Although some Kilosa villagers felt they benefitted throughout the project through things such as trial payments, per diems (for attending meetings) and training, sacrifices were made in anticipation of future benefits via carbon payments. These sacrifices included people being relocated from farms and being permitted from continuing with certain livelihood practices in the VLFRs. It is also worth noting at this point, that by not piloting the carbon payments, the NGO in Rungwe were able to avoid many of the negative impacts of the hype and disappointment cycle, with one village leader reflecting that 'if everybody [in the village] would have known about [potential carbon payments] it would have been a problem. It's a good thing they didn't know this'. 23 This is not to say that the approach taken by the Rungwe NGO was without issue, in fact a number of concerns were identified by the villagers, including in relation to low levels of project participation and a range of concerns were identified by villagers. Nonetheless, within this analysis of expectations, the case of Rungwe provides an interesting contrast in which donor funds were used largely to expand existing activity. This comparison brings up issues of responsibility and accountability for expectations and disappointment, which will be discussed in more detail in Section 5.6.
However, hype and disappointment cycles can have further impacts, including the apportioning of blame to certain actors, damaged credibility and resistance to future innovations (Brown, 2003;Sung & Hopkins, 2006;Van Lente, 1993). At the national level actors directed their disappointment in a number of ways. Some blamed the fact that the donors 'walked away'. 24 Other actors blame the 'top-down' approach of REDD+ 'convincing people to do what they want them to do', 25 which for some actors included criticism of a lack of resources allocated to district and local government. A number of national actors, including national government officials, NGOs and academics were critical of the use of pilots in the future and some reflected that maybe the NGO implementing the Rungwe project took the right approach, avoiding expectations and using the money to continue existing work. However Borup et al. (2006) argue that criticism and disillusionment following hype and disappointment can quickly be pushed aside in the face of a new innovation and new hype. In the future-oriented world of conservation and development where actors have to compete for scarce resources, the hype of new programs that come with promises of multiplewin solutions and donor support may override the critical learning from the REDD+ pilot process (Mosse, 2005;Redford et al., 2013).
In Kilosa, the longer-term impacts of the hype and disappointment cycle of the REDD+ pilot projects were still not fully evident at the time of data collection. However, actor narratives indicated that the experience of the REDD+ pilots had not led to resistance to future projects, although there was a desire for future projects to be done differently among those who felt disappointed. As one K1 farmer still in dispute with village leaders over farming in the VLFR explained: 'It's not that we [would] refuse the projects, we [would] accept the projects to come. They should come but we must make sure we've sat down and plan for that project, together with the village government.' 26 This perspective may to some extent be a product of the fact that this was the first large international forest conservation project implemented with these villages. In situations where multiple previous projects have come and gone, and more hype and disappointment experienced, more evidence of resistance can be found (Leach & Scoones, 2015;West, 2006). In relation to electronic technology, Konrad (2006) finds that hype and disappointment can lead to damage to the credibility and legitimacy of innovators. In Kilosa, those most disappointed with the project largely blamed local leaders for project failures, as opposed to the implementing NGO. As such it appears that the credibility of the NGO remains intact, which may be due to the fact that they have maintained a presence in the villages and have introduced a new sustainable charcoal project. NGOs face a significant challenge in situations such as this, maintaining their credibility and legitimacy among village level actors while engaging with ever more uncertain global mechanisms such as REDD+ (Dressler, 2017).

The social dynamics and 'management' of expectations
The way that different actors and actor groups framed and understood expectations in relation to the REDD+ pilot projects depended on their own individual circumstances, and factors such as their past experiences, social context, personal values and the different ways in which they view or know the world (Leach & Scoones, 2015;Sung & Hopkins, 2006). In Kilosa the villager narratives suggested that the fact that the REDD+ pilot project was the first major donor-funded project, and as such an unknown entity, influenced perceptions. This can be evidenced through the high expectations that surfaced for actors during the early project stages, as well as the fear that the 'country had been sold '. 27 It could also be the case that in Rungwe the long history of conservation and development interventions and the longstanding relationship between the villages and the NGO contributed to the low impact and expectations there. Actors in Kilosa also framed project concepts with their own ways of knowing, for example the framing of the process of 'harvesting of carbon air'. 28 This framing of carbon as tangible and sellable subsequently influenced ongoing expectations in relation to payments.
The experience that people had during the project itself also influenced the way they framed expectations and disappointment (Brown & Michael, 2003). As we have previously identified, those most closely involved with the project put less emphasis on the lack of continued trial payments and as such experienced less disappointment. Conversely, those less involved in the livelihood projects or those who had experienced negative personal impacts focused more on the unfulfilled promises of the project. In K2, where the NGO had brought in a new sustainable charcoal project, MKUHUMI was framed by some as continuing under a different guise. This in turn impacted expectations, with one villager explaining that in the future he expected that 'this MKUHUMI will just be changing its name' 29 but would keep going. We can therefore see that expectations are continually influencing and being influenced by social interactions and experiences (Konrad, 2006) and as a product of the economy of expectations (Dressler, 2017). This evidence also shows that when reflecting on expectations and disappointment, actors re-frame their experience in light of what actually happened and in light of their personal experience (Brown & Michael, 2003). Practitioners need, therefore, to be mindful that no matter how they frame pilot projects to communities, the social dynamics and economy of expectations will be unpredictable, making expectations unmanageable once raised (Weszkalnys, 2008). NGO practitioners involved in the Kilosa pilot project described how they tried to manage expectations around REDD+ and carbon payments as the project developed. One NGO practitioner described how they tried to focus on 'the conservation parts and other cobenefits that they received', but noted that despite these efforts '...we really could not control that [expectations] -there were a few members who... really had high expectations.' 30 This further emphasizes the aforementioned trade-off between raising awareness and raising expectations. Brown and Michael (2003) argue that actors with close proximity to knowledge production have higher levels of uncertainty and lower expectations, while actors further away from knowledge production have low uncertainty and high expectations. Those closest to the production of REDD+ and pilot project knowledge reflected on low expectations and high uncertainty, which as we have discussed was cited by some as rationale for piloting, and low expectations. This included actors with international links, including from the UN, the donor (embassy), international NGOs, universities and consultancies. Among government officials at national and district scales, who can be seen to be further from the production of knowledge around REDD+, there is some evidence of higher expectations and lower uncertainty, particularly in relation to continuation post-pilot: However despite this, members of the Task Force were concerned about the speed at which the pilot projects were unfolding, challenging the testing of benefits at the village level when there was uncertainty as to whether there would be REDD+ benefits long-term.

Knowledge, uncertainty and expectations
Local NGO project implementers reflected on their personal uncertainties and described how they tried to communicate them to villagers: 'Even us ourselves we were not sure about this carbon credit. We were explaining to [the villagers] that this is something new so we are not sure. Even myself, I have been asking that hmmm where will this lot of money come from. Where?' 33 However as we have discussed, expectations rose quickly among villagers, despite attempts to manage them. As such, the Kilosa villagers, who were furthest from the production of knowledge, had the highest expectations and the lowest levels of uncertainty, thus aligning with the pattern identified by Brown (2003). West (2006) finds that villagers engage with all conservation and development projects on the understanding that they are entering into long-term, reciprocal social relationships with practitioners. Even though the villagers in Kilosa were told that the project was time-limited, it appears that they did indeed perceive their involvement as longer term, which it is likely will also be influenced by the nature of the REDD+ mechanism and its emphasis on future benefits (Dressler, 2017).
In light of the analysis in this paper, issues of accountability in relation to expectations in conservation and development policy and practice are raised. This is particularly salient in relation to transparency, responsiveness and liability, which are defined as three of the five dimensions of accountability (Koppell, 2005). In this context, transparency can be seen to be concerned with how project uncertainties (held by those producing the knowledge and leading conservation and development programs) can be better communicated to those furthest away from knowledge production, such as Kilosa villagers (Brown, 2003). This would require a significantly increased level of caution at the start of projects to reduce hype. The challenge for this, however, is that the 'success' of projects relies on hype, raising expectations and the enrolment of actors into communities of promise (Brown, 2003;Büscher, 2014;Mosse, 2005). This again speaks to the need for the tradeoff between piloting and raising expectations to be seriously considered by conservation and development policy-makers and practitioners, which includes challenging the discourse of 'needing' to pilot new program ideas.
Responsiveness refers to whether stakeholder expectations have been met, and liability is concerned with whether consequences were faced by implementing organisations for any shortfalls (Koppell, 2005). The NGO implementing the Kilosa pilot project (along with other implementing NGOs, as reported by their national representatives within this research), were responsive to the expectations of the donor by fully testing the REDD+ mechanism and delivering on project objectives. It can also be argued that the NGOs were liable in relation to their donor accountability, 27  which included analysis of their performance in relation to project objectives. As such, the NGO in Rungwe were criticised for not fully testing the REDD+ mechanism and were challenged by the donors for their choice to use the money to continue with their 'core business' 34 instead of pushing the REDD+ agenda; a choice the NGO took partly due to fears around village-level expectations. The fact that the NGOs have continued to work with communities after the REDD+ pilots through new funding and projects demonstrates their responsiveness to the needs and expectations of the villagers. However, broader accountability for the fact that the REDD+ pilot projects did not meet villager expectations has not been taken, or formally discussed, by the donors and policy-makers who have driven the REDD+ agenda internationally and in Tanzania. Similarly, liability for the disappointment for unfulfilled expectations at the village level has not been taken. This therefore highlights the need for more accountability to be taken by those closest to the production of knowledge for the (unfulfilled) expectations of those furthest away from knowledge production, such as villagers. This echoes wider calls for a shift in how accountability is dealt with in conservation and development policy and practice (cf. Brechin, Wilshusen, Fortwangler, & West, 2002;Campese, 2009;Jepson, 2005).

Conclusions
By applying concepts from the sociology of expectations to the case study of REDD+ pilot projects in Tanzania, we have contributed new insights into the dynamics of expectations in the context of conservation and development pilot projects. By exploring expectations in this way we have also contributed new insights into the understanding of pilot projects, and interventions more broadly, as agents and outcomes of social change. The case of REDD+ in Tanzania demonstrates the important role of hyper expectations in new international conservation and development programs, driving and being driven by a desire for new multiplewin approaches to forest governance, a perceived need for speed, and high estimations of future success. These expectations can be seen as being highly performative, mobilising resources and driving communities of promise among conservation and development professionals. We therefore add insights into the growing critical discussion of conservation fads (Fletcher et al., 2016;Redford et al., 2013), by unpacking the performative role of expectations in this process. High levels of uncertainty existed among those closest to the production of knowledge, yet instead of promoting caution, this uncertainty contributed to a perceived urgency to test the mechanism and drove the 'need' to implement pilot projects. This process can be seen to be a product of what Lund et al. (2016, p. 133) refer to as conservation and development 'logic' that 'continuously produces and feeds off the development and testing of new policy models.' Through exploration of two very different REDD+ pilot projects, we have identified a trade-off between fully testing pilot projects and raising awareness, and raising expectations at the village level. Comparing these two NGOs is not done with the intention of judging or evaluating the NGO approaches or the projects themselves, rather it provides an interesting comparison in relation to expectations. In Kilosa, where the NGO achieved high awareness and high participation in the pilot project, we have shown how an economy of expectations developed (Dressler, 2017). Expectations were raised through project activity, including through wellintentioned activities such as FPIC and testing benefit-sharing mechanisms. Expectations then became a forceful presence (Van Lente, 2012), leading to significant social change, including people being relocated from farms. Expectations interact and are medi-ated by local realities, and so are difficult to manage once raised. A hype and disappointment cycle was identified in Kilosa and expectations have continued to impact villagers after the pilot project finished and the international and national actors have moved on. Conversely in Rungwe, where the market-based aspects of REDD+ were not tested (due to uncertainty about the future of the mechanism and concerns about expectations), there were few expectations and so little evidence of disappointment. Perhaps these two cases reflect the different approaches that the two NGOs take in relation to the challenge of maintaining legitimacy with village level actors while engaging with ever more uncertain international pograms in the competition for funding (Dressler, 2017).
Our findings therefore highlight some core issues for conservation and development and support calls for more critical reflection of how conservation is pursued, particularly in relation to how new international programs such as REDD+ are managed (Fletcher et al., 2016;Lund et al., 2016). Expectation and disappointment cycles can be conceptualized as an unintended consequence of piloting new international conservation and development programs, particularly in relation to future-oriented, market-based programs such as REDD+ (Dressler, 2017;Igoe & Brockington, 2007). However, although unintended, expectations are inevitable, which the trade-off identified in this research demonstrates. The negative outcomes of hype and disappointment cycles are asymmetric; produced by those closest to the production of knowledge and yet impacting those furthest away from knowledge production the most (Brown, 2003). This is particularly salient in relation to pilot projects, which are framed as a short-term test by international actors but seen by local actors as being the start of a longerterm, reciprocal relationship (West, 2006). Accountability for expectations is therefore needed in conservation policy and practice, particularly on the part of those closest to the production of knowledge, such as policy-makers and donors. This includes the need for more transparency around uncertainty from the start, more responsiveness to villager expectations and liability being taken for unfulfilled expectations. To this end, we challenge the discourse of 'needing' to pilot, which prioritizes awareness, impact and innovation without fully considering the potential negative impact of unfulfilled expectations.

Declaration of interest
None.