‘Developmental nationalism?’ Political trust and the politics of large-scale land investment in Magufuli's Tanzania

ABSTRACT Research on large-scale land investments (LSLIs) can provide valuable insights into the support for developmental nationalism in Tanzania today. ‘Developmental nationalism’ is ‘a creative variant of liberation’, which purports to make ‘Tanzania great again’. The nationalist turn of late President Magufuli was grounded in political ideology and the selective history of the past that swept him to power. However, there is limited research on how political practice around land investments contribute to trust and support for public institutions. This paper makes two key contributions to scholarship on the political economy of LSLIs. First, we examine the messy politics of LSLIs, the failures in design and implementation, and the rise in local support for developmental nationalism in two rural settings in Tanzania. Second, using Latent Class Analysis (LCA), we identify distinct groups of individuals based on their trust in the President, the ruling party (CCM), the Tanzania Investment Centre (TIC) and support for LSLIs. We define political trust as ‘an evaluative orientation towards an institution or government, based on people's normative expectations’.

Research on large-scale land investments (LSLIs) shows that land investments are unquestionably political, particularly within statist land tenure regimes in Africa. 1 Schlimmer demonstrates how the 'land grab' discourse became a talking point during the 2015 presidential elections in Tanzania. 2 Cliffe et al. explain how the Fast Track Land Reform Programme (FTLRP) in the early 2000s shaped and reformed Zimbabwe's political and economic environment. 3 In countries where 'land laws are centralised, and land is vested in the president as trustee', the direct involvement of government agents and officials in land allocation and dispute resolution can foster the politicisation of LSLIs. 4 However, there is limited research on how political practice around land investments contributes to the trust and support for public institutions in the Global South. The direct involvement of state agents and officials in land transactions can transform land investments into a tool for politicians to garner political support. 5 For example, Bélair describes the land transfer process in Tanzania as 'political', with rural communities frequently dispossessed of land and the 'public interest' used to justify expropriation. 6 This paper makes two key contributions to research on the politics of LSLIs. First, we examine the politics of LSLIs, the failures in design and implementation, and the rise in support for 'developmental nationalism' in two rural communities in the Ruvuma region of Tanzania. Second, using household data, we employ a latent class analysis (LCA) to identify groups distinguished based on their trust in the President, the Tanzania Investment Centre (TIC), in the ruling political party Chama cha Mapinduzi (CCM) and their support for LSLIs. To determine if there is a significant difference between the classes identified in the LCA, we employ McNemar's dependent Z-test. We define political trust as 'an evaluative orientation towards an institution or government based on people's normative expectations'. 7 Like Paget, we argue that as the insurgent leader of CCM, late President John P. Magufuli instrumentalised developmental nationalism as a political tool to increase political support among rural communities, already disillusioned by unfulfilled promises within LSLI schemes. 8 Paget describes this developmental nationalism as 'a creative variant of liberation', which purports to make 'Tanzania great again '. 9 By 'liberation', Paget describes a form of progressive restorationist ideology under Magufuli's regime, which breaks away from an 'old order' laissez-faire economy that has disenfranchised the majority of smallholder farmers for over 20 years. 10 Developmental nationalism is versatile, mutable and legitimised Magufuli's authoritarian rule in the eyes of his supporters. 11 The nationalist turn of Magufuli was grounded in political ideology and a selective history of the past. For one, an article in The Economist describes Magufuli's leadership style as having had 'a whiff of African Socialism'. 12 Cheeseman argues that Magufuli's leadership did not represent an actual break with the past but can be understood in the context of the country's return to a more statist economic approach, last experienced under Nyerere. 13 For Magufuli to legitimise his decisions, he drew on the revered legacy of Nyerere, which led some authors to qualify his actions as 'post-socialism'. 14 How did this manifest? The President publicly attacked foreign and local private investors as corrupt and succeeded in presenting himself as fighting for the poor against a corrupt business and political elite. 15 The high levels of distrust in political institutions, which were evident in the run-up to the 2015 election, 16 were demonstrated by ordinary people who perceive public institutions like the TIC as favouring powerful and corrupt investors against local community members. 17 For analytical purposes, we present the TIC as representing 'old order' laissez-faire institutions that advocate for a neoliberal and open economy in favour of LSLIs. 18 On the other hand, Magufuli introduced a 'business unusual' approach to investments which was supposedly aimed at correcting the shortfalls in Tanzania's relationship with investors, 19 a form of developmental nationalism.
Institutions like the TIC have encouraged LSLIs to increase foreign direct investments (FDI), farmers' incomes, technology transfer, and job creation. 20 However, local communities, disillusioned by the broken promises of these LSLIs, 21 expressed a kind of wistfulness and nostalgia that can apparently only be brought back by a restorationist form of developmental nationalism, as expressed by Magufuli. 22 The negative impacts of LSLIs on the rights of local communities, 23 and the absence of adequate political tools oriented to holding investors accountable for the promises made to local people, 24 place public institutions and citizens in a constant struggle for political legitimacy and support. 25 Considering the analytical challenges in measuring 'political trust', 26 our paper accepts certain analytical assumptions. We do not generalise our findings. However, by using two rural communities, one where a LSLI project has been implemented and another where the investment project has not been implemented, our research provides valuable insights into the politics of LSLIs in rural communities in Tanzania. In the next section, we elaborate on the analytical and data collection techniques used in measuring political trust as a latent variable for political support.
Measuring 'political trust' as a latent variable for political support Challenges abound in measuring political trust as a latent variable for government support within social science research. 27 In looking at the trends in political trust within new and stable democracies, Catterberg and Moreno also define political trust as 'citizen's confidence in political institutions'. 28 Our analysis focuses on smallholders' evaluative orientation towards institutions based on their expectations of the promises made within LSLI schemes. However, trust is a subjective and normatively charged concept. 29 What constitutes a trustworthy institution is likely to differ among citizens within different cultural and governance regime contexts.
To mitigate contextual misinterpretations in our study, we focus on two rural communities in the same local administrative region. Fieldwork was carried out from May to September 2018, following a three-week-long pilot study in December 2017. Previous visits by the first author to these communities in 2014 as part of his MSc research proved valuable in understanding local dynamics and for carrying out research 'where context specificity is essential'. 30 Like James Scott, we assume that our sample community is fashioned from the same everyday realities, and their divergent experiences and interests form part of the same community of discourse and practice. 31 Our political trust analysis within the same regional and socio-cultural context thus minimises 'measurement errors', which may occur when such studies are done across cultural and regime settings. 32 Village 1 is 50 miles away from Village 2 and falls within the same ethnolinguistic, cultural and political economy context.
The choice of Village 1 is attributed to its proximity to a functioning 1,999 ha largescale farm, whilst Village 2 is adjacent to a failed investment project of about 20,000 ha. Foreign and local private investors have made investments in Villages 1 and 2, respectively. Both villages are located within the Southern Agricultural Growth Corridor of Tanzania (SAGCOT), a public-private partnership scheme launched in May 2010 by President Jakaya Kikwete. 33 The SAGCOT covers 350,000 ha of 'profitable land' and aims to produce 680,000 tonnes of field crops (maise, tea, soya and wheat), 630,000 tonnes of rice, 4.4 million tonnes of sugarcane, 3,500 tonnes of red meat and 32,000 tonnes of high-value fruits by 2030. 34

Data collection
We used different techniques to collect data on the perception of local communities. First, we relied on key informant interviews and focus group discussions (FGDs). Key informant interviews (n = 18) were conducted using unstructured interviews with village elders, representatives of farmers' based organisations, academia, state and local government officials. The key informants were selected through a snowballing technique. The FGDs (n = 7) helped us understand how local community members perceived land investment policy design and implementation and why they consequently challenged the performance of the LSLI schemes in their communities. FGD participants were selected from each village hamlet to allow for fair and equal geographical representation across the villages (see Maps 1 and 2). Where possible, an elder with informed knowledge of the village history was preferred for the FGD.
In Village 1, FGDs were organised with the 14 participants equally split into two groups of seven males and seven females. However, in Village 2, where another 14 participants were selected, the FGD was limited to one session (of 2 hours) with male and female participants because of the abrupt termination of the first author's research permit. 35 Additionally, grey data and documents such as letters between villagers and government officials, minutes of village meetings and land use maps were obtained from village members and subjected to content and thematic analysis. The documents were also used as prompts during FGDs.
Second, with the help of two research assistants (one male and one female), we administered a standard Swahili-translated 'trust in government' survey to households in both communities. Questions were translated beforehand from English to Swahili to minimise inconsistencies that can occur when questions are translated independently by interviewers. Table 1 shows a statistical description of interviewees by age, gender and educational level. An average of three adults per household was interviewed in 112 households in Village 1 and 75 households in Village 2. Thus, a total of 374 interviewees responded to the survey questionnaire. Where possible, male and female adults in a household were interviewed. Interviewees within the same household were interviewed separately to minimise the influence that other housemates or partner(s) had on an interviewee's responses. Table 2 presents the trust variables and how they were coded for analysis. These standard trust in government variables have been used in other studies on political trust. 36 To link the LSLIs and the public institutions, we used the 'support LSLIs' variable. Authors like Suh et al., have used the LCA approach to classify individuals based on their trust in public and non-public (or private) institutions. 37 Map 2. Land use map of Village 1 including distribution of household survey by hamlets. These rigorous data collection techniques helped minimise measurement errors in our political trust analysis and develop a clear narrative of LSLI schemes in these communities. In the next section, we start by building an argumentative groundwork of the politics of LSLIs and how it may contribute to the support of developmental nationalism under Magufuli in rural Tanzania.

How local politics becomes entangled in the politics of large-scale land investments
In this section, we present empirical data on how local political economies become entangled in the politics of LSLIs and highlight how the developmentalist perception of the state officials as being 'responsible for, but not to, citizens' creates animosity between public institutions and local communities. 38 In terms of periodisation, we focus on the discourse from the late 1980s, when the first half of the land in Village 1 was acquired. This is a period which coincided with the introduction of neoliberal policies in Tanzania. We also highlight the formation of the TIC following the enactment of the 1997 Tanzania Investment Act and look at the broken promises and adverse outcomes that accompanied LSLIs projects under the Kikwete regime (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015). Understanding how and why people act in politics and why there is support for developmental nationalism in these local communities demands that history be understood from the standpoint of its agents. 39 The acquisition of the 1,999 ha piece of land in Village 1 happened in two phases. The acquisition of the first half of the land, 404 ha (see Map 1), occurred in 1984 when Tanzania transitioned from a socialist regime to a neoliberal political economy. 40 The second half comprising 1,595 ha was acquired in 2011 when the Kikwete administration launched the SAGCOT scheme. 41 The 20,000 ha of land in Village 2 (see Map 2) was acquired in 2010. These periods were marked by an active invitation to and support for private investors in agriculture. 42 Nelson et al. regard the increase in the competition for land by foreign and local investors in Tanzania as an outcome of political and economic changes, which themselves led to the liberalisation of the Tanzanian economy 1980s. 43 Ngoitiko et al. argue that the reforms in the 1980s gave rise to an unprecedented period of land grabbing in Tanzania. 44 These debates, including those of the 2015 electoral campaign, are thus not new. The re-emergence of the discourse in the 2015 general elections merely extended and reproduced ongoing debates over land and investments in Tanzania. 45 What follows is a review of the political and economic changes and how these changes contributed to LSLIs at the local level. Political competition and the introduction of neo-liberal policies The 'shift to an open market economy in the 1980s meant that Tanzania now had a government which supported private investments'. 46 To obtain structural adjustment loans, President Ali Hassan Mwinyi (1985)(1986)(1987)(1988)(1989)(1990)(1991)(1992)(1993)(1994)(1995) adjusted the developmental ambitions of Tanzania to the conditionalities of foreign financiers, including the International Monetary Fund (IMF) and the World Bank. 47 These conditions included: (1) the devaluation of the Tanzanian shilling; 48 (2) changes in the political structure, introduced by the 1991 Zanzibar Declaration and multipartyism; 49 and (3) changes in the investment regulatory framework, including investment and private property laws. 50 By 1991, state subsidies provided to farmers were suspended under the Economic Recovery Programme (ERP). 51 Introduced in 1986, the ERP encouraged private sector engagement in agriculture and reduced state interference in the market. 52 As part of the ERP, the Civil Society Reform Programme of 1991 was implemented in 1993, 53 and publicly owned enterprises were liquidated and acquired mainly by the business elite. 54 These economic and political changes led to widespread redundancies in the public service and increasing inequality between the business elite and rural sections of society. 55 Consequently, political debates about the expansion of the economy became entangled in what Aminzade refers to as the 'indigenisation debates'. 56 For example, in the mid-1990s, opposition party leaders like Rev. Christopher Mtikila of the Democratic Party described Tanzanians of Asian and Arab origin as thieves and looters of the country's wealth (gabacholis) at the expense of indigenous Africans or the downtrodden people (mkombozi wa walahoi). 57 The economically better-off Tanzanians, mainly of Asian origin, were seen as better able to acquire assets while the impoverished (African) majority became further economically marginalised in their ancestral land. 58 It is argued that Tanzanians of Asian origin were economically privileged to acquire significant private assets and became the new proprietors of many state-owned enterprises in the wake of economic liberalisation. 59 In a seemingly strategic move, ruling political party leaders like Idi Simba began substituting the indigenisation rhetoric as a struggle against foreign interference. 60 Here, indigenisation drew a difference between citizens and foreigners. 61 The indigenisation rhetoric became a populist tool to gain political support by advocating for the protection of the nation's economy and culture from the threat of foreign domination. 62 The rising inequality legitimised the indigenisation debates and remained a dominant issue in the newly established multiparty system. 63 Mwenye Busara (not his real name), a 70-year-old key informant in Village 1, remembers the socio-political changes during this period. Mr Merali, of Asian origin, benefited from the newly liberalised economy in the 1980s and was able to acquire the piece of land in this Ruvuma village. Mwenye Busara and a group of the village elders acquiesced to the transfer of 404 ha of land to Mr Merali in 1984. 'We were young when our land was taken. I was 36 years old. He [Merali] told us that he would bring jobs and develop our village '. 64 Changes to investment and land laws for increased private sector engagement By 1993, recognising the need for legal and institutional reforms to 'facilitate an enabling environment for enhanced private sector participation', the government announced the establishment of the Legal Sector Task Force led by Mark Bomani. 65 According to the Bomani report, 'market-oriented economic reforms could not be implemented in the absence of a "sophisticated", legal or regulatory framework capable of meeting the exacting challenges of a "modern" market economy'. 66 Consequently, the 1990 Investment Promotion and Protection Act, which protected the local industry, was scrapped and formed the basis for enacting the 1997 Investment Act. 67 To appease advocates calling for the indigenisation of the economy, vocal proponent and businessman, Idi Simba, was appointed Minister of Commerce and Industry in 1999. 68 After his appointment, opposition party parliamentarians joined 'CCM backbenchers to support the National Employment Promotion Service Act of 1999', which introduced measures designed to require foreign investors to enter into business partnerships with Tanzanian citizens. 69 This encouraged the link between foreign investors and local business people, in what authors like Shivji have described as compradorialism. 70 The new 1997 Tanzania Investment Act led to the creation of the TIC as a one-stopshop to facilitate the acquisition of land and investment certificates for investors. 71 These legislative and institutional changes opened up all sectors of the economy for investment. Significantly, the 1997 Investment Act abolished the protection of local industries that was provisioned in the 1990 Investment Promotion and Protection Act. 'The best investment law this country has ever had, and which gave locals a fair opportunity for investments, was the 1990 Investment Promotion Act. Surprisingly, Tanzanians spearheaded the change of the law … they were just a group of comprador guys'. 72 To authors like Zoomers and Nelson et al., the interest in land investments was a consequence of 'the liberalisation of land markets, which became a major policy goal in the 1990s'. 73 The liberalisation of the economy responded and contributed to the commodification of land and other natural resources.
Additionally, the absence of robust regulatory and oversight institutions allowed corrupt politicians and institutions to engage in rent-seeking with devastating consequences for the economy. 74 For example, the Local Government Reform Act of 1998, which was introduced to decentralise decision-making processes within government, failed to yield a veritable devolution of power due to political and institutional constraints. 75 Significantly though, changes to land laws in 1999 introduced the oversight mechanism needing the consent of the village assembly, i.e. all village adults above 18 years of age, for any transfer of land to a potential investor. However, the 1999 land laws left the administration of village land transactions under district authorities. 76 In principle, the 1999 Village Land Act of Tanzania provides for the separation of power between the Village Assembly and the Village Council, as the bases for villagelevel governance. 77 However, local elites, including government officials, continued to perceive village community members as the objects of development and recipients of orders from above rather than co-partners in decision-making processes. 78 The land reforms in 1999 were criticised as 'being too focussed on enhancing economic growth', with the aim to enhance the productivity and profitability of the agricultural sector as stated in the Agriculture Sector Development Strategy of 2001. 79 For instance, the Strategic Plan for Implementation of Land Laws (SPILL) criticised pastoralists and smallholder farmers for not contributing to agricultural productivity and economic growth. 80 With the transition to multiparty democracy in 1995, corruption became a significant driver of change in Tanzania. 81 Continued inefficiencies in institutional and regulatory frameworks in the wake of market-led investments affected the functioning of state institutions. The implementation of the land law reforms remained problematic because of the inadequate resources and limited capacity at the local and national level, including the Ministry of Lands. 82 The falsification of VAT receipts at the Tanzania Revenue Authority to the tune of 15 million dollars per annum in 2003 83 and increasing tax exemption cases of approximately 178 million dollars in 2001 contributed to losses in tax revenue by the government. 84 By 2005, when Kikwete became President, the contradictions between the ideas of the ruling class and the lived experiences of common Tanzanians had become profound. 85 A review of the Tanzanian Development Vision (TDV) 2025 by Mashindano et al. (2011) showed that the rate of economic growth that had followed economic liberalisation did not correlate with poverty reduction and improved livelihoods for ordinary Tanzanians. 86 Many investment projects had failed to materialise into profitable economic opportunities for smallholder farmers. 87 For example, the land which Mr Merali acquired in Village 1 remained vacant till 2011. 'We accept that we gave the land to Merali. We gave [it to] him because he promised to bring jobs and develop our village. So, we did not go to that land since it was no longer our land'. 88 The increasing poverty levels led the government to allocate appropriate resources for accelerated sustainable and inclusive economic growth. The outcome was an increase in LSLI schemes and a focus on the commercialisation of agriculture, which, according to Kelsall, became populist vehicles for promises during election campaigns. 89

Broken promises
Following the increasing focus on LSLIs under President Kikwete's administration and the global land rush after the 2007/2008 food crisis, 90 legal and institutional processes were often undermined, making customarily held land vulnerable to loss and dispossession. 91 By 2009, an estimated 4 million hectares of land was requested from the Tanzanian government, through the TIC, with about 640,000 ha having been formally allocated to investors. 92 In August 2009, President Kikwete launched the Kilimo Kwanza -'Agriculture First' strategy as a national resolve to accelerate agriculture transformation in Tanzania by 'providing incentives to attract more agricultural investors'. 93 The Kilimo Kwanza strategy was Kikwete's plan for how commercialised agriculture would contribute to economic growth. 94 As part of the Kilimo Kwanza, the SAGCOT was launched in May 2010. 95 The SAGCOT encompasses Dar es Salaam, Morogoro, Iringa, Mbeya, Ruvuma, Njombe, Rukwa & Katavi, and is supported by the G8 ′ s (now G7) New Alliance for Food Security and Nutrition Strategy. 96 Transnational corporations (TNCs) like Yara, Monsanto, Bayer Crop Science, Syngenta, Unilever, the World Bank, and the US's Feed the Future strongly supported the initiative. 97 While an objective of most governments in developing countries is to attract muchneeded capital through FDI, the benefits of the commercialisation of agriculture remain debatable in these countries, and institutional performance can influence support for such policy interventions in the short and longer term. 98 The launch of the SAGCOT introduced the second phase of the land acquisition in Village 1, only this time, without the knowledge and consent of village members. 'We were shocked when we received a letter from the village executive officer (VEO), restricting us from accessing the land on the left side of the road'. 99 The son of the now-deceased Merali returned to the village in 2011 to claim his father's land (404 ha), including an additional 1,595 ha of land on the left-hand side of the road, the 'Luchini area' (see Map 1).
Land laws in Tanzania remain highly centralised, such that the President can acquire any land for any public purpose. 100 Land transfer processes remained political, and rural communities have been dispossessed of land frequently, with the 'public interest' used to justify land expropriation. 101 It came as a surprise to Mwenye Busara and his fellow village members that this investor claimed an additional 1,595 ha.
When the current investor came, we were grown up. We witnessed a massive conflict between the investor and us [sic]. Of course, as villagers, we did not offer our consent (to the 1,595 ha), but we had to give [it] up because all our efforts yielded no results. We struggled by writing several letters to the authorities to bring back our land, especially the land on the left side of the Road [the Luchini area]. The poison of our grabbed land is still in our hearts. 102 Similarly, the land in Village 2 was acquired in 2010, and village members have since been barred from accessing the piece of land, which also functioned as their cemetery. Indeed, several LSLI deals in Tanzania have been reported to disregard village members in support of investors who possessed capital for investment and development. 103 A research participant described the investor in Village 2 as 'just a cunning man'. That is why he got the land. 'He took advantage of our illiteracy. None of his promises was fulfilled!' FGD participants said the only useful thing for them when the land was transferred were the 'promises he made to us'. 104 In many cases, like in Village 1 and 2, contestation over access to land is often spurred by an investment and development agenda which undermines rural communities' control over their land and other productive resources. 105 The focus on the commercialisation of agriculture serves as a vehicle for local political elites to pursue a developmentalist agenda and see themselves as empowered to rule for the good of society and treat village communities as marginally competent political agents. 106 The political imagination of the state elite is often expressed in deeply paternalistic and hierarchical terms, with their 'knowledge' and 'modern' agency placed alongside the 'ignorance' and 'backwardness' of village members. 107 Unsurprisingly, rural communities like Village 1 and 2 become political hotspots to boost popular support for public policies towards poverty reduction and income generation through the commercialisation of agriculture. It explains the visit of President Kikwete to Village 1 in 2014, where he raised the expectations of community members by saying that he had directed government institutions to increase the pace of poverty reduction in the country, which could be done through investments such as the newly established large-scale farm in the village. 108 While conducting research in Village 1, the first author overheard a local police officer remark that 'village members have a low IQ', and that this research was introducing ideas that could cause them to riot against the investor and the government. 109 However weak the material evidence of such claims, they find resonance in wider Tanzanian society by linking LSLIs to poverty reduction, employment, and local communities' rights. 110 For example, in August 2011, during a parliamentary session, MPs such as John Cheyo of the United Democratic Party questioned the country's land laws for 'allowing foreign investors to hold large chunks of land while natives are marginalised'. 111 There is a general perception that LSLIs are often linked to a corrupt political elite and foreign investors at the expense of disadvantaged local communities. 112 The increase in land conflicts between investors and local communities disenfranchises local communities from the 'benefits' of LSLI in various parts of Tanzania. 113 The dire food insecurity experiences of research participants in Village 1 and 2 are documented in a separate paper. 114 For example, when Mwenye Busara was interviewed, he was weaving a basket from raffia-like twines, which he planned to exchange for 4 kg of maize per basket, maize being the staple in Ruvuma. Over 90% of research participants from the household survey in Village 1 and 2 described maize-based stiff porridge (ugali) as the staple within their households, often eaten on a daily basis. Ugali is usually accompanied by vegetables and fish. Halfway into the interview, Busara requested some food from one of the research assistants. Busara was hungry, and his daughter who is also his care, with whom he lives in a tiny mud hut, had left for work on the large-scale farm. Over 47% of interviewees reported experiences of being severely food insecure. 115 This represents the socio-economic situation in Village 1, where many of its members are labourers in what is now a large agribusiness.
During the FGD in Village 1, community members seemed to have given up on the struggles to defend their land. 'This is a massive investment, and you cannot fight it with the money from begging [sic]. For our chairman to go to town to follow up [these issues], he has to borrow the transport fare from various shops within the village'. 116 The failures of investment and economic policies in Tanzania legitimise support for the nationalist turn in Tanzania today. Poncian argues that the crisis of legitimacy of the neoliberal orthodoxy is leading to a rise of nationalism in many countries, including Tanzania. 117 During this research, it became clear that village members associated anything to do with investments as the failures of the TIC, which explains the low level of trust in 'old order' laissez-fair institutions vis-à-vis Magufuli's new developmental nationalism approach. 'When a leader [Magufuli] comes as a sovereign state and rejects things that do not align with the state's sovereignty, he must correct them. Correct those things so that Tanzanian natural resources are exploited to help Tanzanians [not foreigners]'. 118 For the majority of smallholders in village communities, Magufuli's actions were viewed as a solution to what Amin describes as the systematic exclusion of smallholders from a capitalist-oriented economy. 119 President Magufuli's election also came against the background of a political climate marred by high-level corruption scandals and distrust in the public institutions in Tanzania. 120 The private interests of state elites, as argued by Nelson et al., became entangled with the development of elite capture and patronage-based politics in Tanzania. 121 For example, Pelizzon and Bekenova provide detailed accounts of some of the high profile scandals, including the Richmond scandal, the Bank of Tanzania's (BoT) External Payment of Arrears (EPA) scandal, the Escrow scandal, and the BoT's twin tower inflation scandal. 122 According to a report by The Monitor, these scandals had robbed the taxpayer of Shs. 527 Billion (229 million USD). 123 Cooksey describes how the Escrow-Richmond scandal, which ran from 1992 and involved several high-ranking politicians and businessmen, including former prime minister Edward Lowassa, cost the Tanzania public vast sums of money only to benefit 'opportunistic rent-seekers'. 124 These scandals, symptomatic of the high level of corruption within Tanzania, contributed to considerable disillusionment and distrust in the legitimacy of the ruling elite.
Additionally, in-fighting between members of the ruling political party CCM led to the resignation of Prime Minister Lowassa in 2008 due to his alleged involvement in the Richmond Scandal. 125 This paved the way for the ascent to the top job of the then Minister of Works (from 2010 to 2015), John P. Magufuli, who had gained a reputation as a no-nonsense Minister, an approach which earned him the nickname 'the bulldozer'. Magufuli's first years in office were dedicated to building political support by introducing popular social policies such as free education for children between the age of 8 and 12 and improved health services for the poor. 126 He also focused on improving the public service and curbing corruption while consolidating power at the helm of a weakened ruling political party. 127 His supporters have heralded this approach to investment policy and regulations under the Magufuli government as attempts to redefine and rebalance Tanzania's relationship with investors. The pragmatism of President Magufuli's fight against corruption and his ostensibly anti-establishment agenda propelled him to the head of a nation and a ruling political party (CCM) that needed fundamental changes in structureprinciples, values and institutionsan image undoubtedly different from the neoliberal reforms introduced in the late 1980s. 128 In a report published in 2015 on the Agriculture Sector Development Strategy of Tanzania, the government of Tanzania has acknowledged that the focus on large commercial farms had made little contribution to the reduction of poverty. 129 Indeed, increasing levels of corruption and the seemingly imminent disintegration of the ruling political party in the run-up to the 2015 elections paved the way for President Magufuli, who would advocate for change, and disrupted the business as usual approach in Tanzania. 130 In the next section, we present the results of our trust in government survey using the LCA.

Trust in institutions as a result of perceptions around LSLI
In this section, we present the results from the LCA using the trust in government survey responses. Respondents were classified using approaches suggested by Vrieze and Suh et al. 131 We ran several models in R using a different number of classes to categorise survey respondents. The analysis showed that our respondents could best be categorised into three classes. The statistical software generates the conditional item response probabilities per variable and the estimated class population share for each class (see Table 3).
The calculated size of each class (n i ) = Estimate class population * total number of observations (n), where i is the class number. From Table 3, we define the various classes as follows: . Class 1: Class 1 respondents have a neutral tendency towards LSLIs and the President, and the TIC. Some 31% of respondents fall in Class 1. Individuals in Class 1 are 96% likely to trust the President, 94% likely to trust CCM, 66% likely to trust opposition parties, 97% likely to trust the TIC and 64% likely to support LSLIs in the community. . Class 2: Individuals in Class 2 are less likely to support LSLI in the community (9%).
This class makes up 44% of the sample population and are 99% likely to trust the President, 97% likely to trust CCM, 35% likely to trust opposition parties, and 0% likely to trust the TIC. . Class 3: Individuals in Class 3 tend to distrust any public institution and LSLIs. Class 3 makes up 25% of the sample and are 40% likely to trust the President, 16% likely to trust CCM, 9% likely to trust opposition parties, 3% likely to trust TIC and 22% likely to support LSLIs in the community.

Difference between classes
Using McNemar's Z-test, 132 we then determined if there is any significant difference between Class 1, where participants are likely to support LSLIs, and Class 2 and 3 where participants are less likely to support LSLIs. In doing so, we set the hypothesis for the differences in classes. Firstly, by inferring from our data, the proportion of those who do not support LSLI in the community (Class 2 and 3) is greater than the proportion of those who support LSLIs (Class 1). Therefore, we set the null hypothesis to test whether those who support LSLI in the rural community (i.e. Class 1) is statistically different to those who do not support LSLIs in the rural community (i.e. Class 2 and 3). There, we execute two hypothesis tests.
Hypothesis test 1:McNemar ′ s Z score = (n 2 -n 1 )/ √ (n 2 + n 1 ) Hypothesis test 2:McNemar ′ s Z score = (n 3 -n 1 )/ √ (n 3 + n 1 ) From Tables 4 and 5, given that Class 1 is greater than or equal to Class 3, but always less than Class 2, we conclude that the population proportion in Class 2, which is the highest proportion of our sample, do not support LSLIs but support the President. This, therefore, supports our claim that despite the promises made within LSLI schemes, community members are disillusioned by the broken promises and would support a president who purports to fight against institutions like the TIC that are linked to these LSLI deals and schemes. Community members are more likely to support the President,  whose approach to investments had elements of developmental nationalism and expresses himself as the leader of the downtrodden. On the other hand, community members are not likely to support LSLIs or trust the TIC, represented here as an 'old order' laissez-faire institution.

Conclusions
The involvement of state agents in land transactions can transform land investments into a tool for politicians to garner political support. These land transactions contribute to questions about legitimacy and government representation within local political economies. Additionally, the vulnerable position of customary land within statist land tenure regimes pits local communities against local government officials who continue to regard them as incompetent political agents. The negative impacts of LSLIs on the rights of local communities, and the absence of adequate political tools oriented to holding investors accountable for the promises made to local people, place public institutions and citizens in a constant struggle for political legitimacy and support. While political trust is a highly normative and subjective concept, the use of a case study approach minimises measurement errors that may occur when measuring political trust across different cultural and regime settings. From the LCA performed it this study, it is evident that there is a difference in support for the president and 'old order' laissezfaire institutions. Results show that rural communities can be categorised into three classes with 31% of rural community members 64% likely support LSLIs. Another 44% are 9% likely to support LSLIs. The third class, 25%, are 22% likely to support LSLIs. In all three classes, community members are more than 95% likely to support the President. While the research does not investigate the successes and failures of developmental nationalism under President Magufuli; it shows that rural communities impacted by LSLIs are likely to support a developmental nationalism approach to investments in Tanzania. Such support or trust is dependent on people's normative expectations towards public institutions linked to LSLIs. For years, local communities have been disillusioned by the promises made within these investment schemes, which are often presented as a means to increase FDI, farmers' incomes, technology transfer, and job creation. The negative impacts of LSLIs on the rights of local communities and the absence of adequate political tools, which are oriented to holding investors accountable for the goals promised to local people, underpin the popular support for the nationalist turn in Tanzania today.
To local communities impacted by LSLIs, Magufuli's actions were viewed as a solution to the systematic exclusion of smallholders from a capitalist-oriented economy. Magufuli instrumentalised this variant form of developmental nationalism as a political tool to increase political support among rural communities. His developmental nationalism 1.53 −1.64 Z score falls within the acceptance zone we fail to reject the null and conclude that the proportion of those who support LSLI, Class 1, is not always less than those Class 3. was versatile and mutable, and legitimised his authoritarian rule in the eyes of his supporters as a fight for the poor against a corrupt political and business elite.