Introduction
In March 2000, land occupations in Zimbabwe intensified, spreading throughout the country and forcing the government to implement the Fast Track Land Reform Programme (FTLRP). The FTLRP significantly altered the agrarian structure of Zimbabwe, largely replacing the predominantly white-owned large-scale commercial farming sector with a broader base of small-scale (A1) and medium- to large-scale (A2) farms. Since then, the debate surrounding the FTLRP has been highly polarised. In recent years the focus has partly shifted from a discussion of processes to an analysis of outcomes, in large part thanks to Ian Scoones' high-profile, but highly controversial and widely criticised, work on livelihoods in Masvingo Province (Scoones et al. 2010).1 However, several sticking points remain. There is, for example, no clear consensus on the nature and character of post-2000 land occupations in Zimbabwe, and widespread misconceptions surrounding the identities of land ‘invaders’ and beneficiaries persist to date. Who are the new smallholders and where did they come from? These questions are simple, but the answers are often complex and nuanced.
The data used in this article were collected in 2011–12 during 18 months of fieldwork in Zimbabwe. Quantitative data were collected using a household survey,2 while qualitative data were collected via in-depth interviews and focus groups with the new smallholders. The sample includes 166 households from four post-2000 resettlement villages. Three of the villages – North Star, Golden Star and Chiraramo – are located in Shamva District in Mashonaland Central and the other – Bita – is in Hwedza in Mashonaland East. The former three are in Natural Region (NR) II and the latter is in NRIII, accounting for two of the five agro-ecological zones of Zimbabwe.3 All four villages belong to the A1 ‘villagised’ model, where households are clustered together in villages, with each plot holder receiving 5 hectares of land with access to communal grazing land and a residential stand. The location of each village is shown in Figure 1.
The article begins by outlining the debate surrounding land occupations and ‘who got what’ after 2000. I argue that the earlier tendency to focus on the role of the state and its use of the land issue to hold onto political power was too narrow because it ignored or downplayed the idea that post-2000 land occupations also represent a continuation of a century-old popular land movement. I also argue that there has been a fixation on the issue of ‘elite’ land grabbing which has overshadowed the wider realities of the FTLRP. I then move to show that the intensification of land occupations after 2000 was the result of extreme frustration among the poor and the landless who – in protest against white privilege and the state's failure to adequately redistribute land – took matters into their own hands. These historical events and processes combined to produce conditions of extreme frustration and shaped the smallholders who now occupy the land. Understanding these processes and the motivations that drove people to occupy land is a vital step towards knowing who the new smallholders are. Lastly, I look at the social background and origins of the new smallholders, present a socio-economic snapshot of the four villages and identify four sub-groups of smallholders. I conclude that the majority of the FTLRP beneficiaries in these villages are ordinary people from less well-off backgrounds.
Debating land occupations and ‘who got what’
Some early responses to the FTLRP drew a clear distinction between past land occupations and those that took place following the constitutional referendum in February 2000. In one case, they are described as ‘mirror images’ of each other (Alexander 2003, 84). Pre-2000 land occupations were seen as ‘mostly spontaneous’, led by land-hungry peasants, unemployed urbanites and war veterans, reflecting a ‘growing discontent with the pace and priorities of the government's land resettlement programme’ (Hammar and Raftopoulos 2003, 9). These ‘squatters’ often faced eviction and arrest at the hands of the Zimbabwe Republic Police (ZRP). Whereas later occupations were seen as a ‘state-led’ and ‘nation-wide campaign of land invasions onto mainly white-owned commercial farms … using unprecedented extremes of violence … . fronted at first by militant war veterans … expand[ing] to include ruling party-trained youth militia and … state security forces’ with the aim of ‘destroy[ing] the world of the white farm’ and ‘punishing and/or eliminating a constituency of white farmers and “their” workers widely viewed to have voted “no” in the referendum, and, by implication, to support the opposition more generally’ (Hammar and Raftopoulos 2003, 10–11). Through this lens, post-referendum land ‘invasions’ could no longer be understood in terms of a popular nationalist land movement demanding more and faster land reforms, but rather an authoritarian state-led political campaign designed to crush opposition and bolster the then ruling party's waning support.
In addition, this literature speaks of ‘explicit allocation of plots along partisan lines, largely excluding opposition supporters at all levels’ and ‘extensive and multiple land grabbing of prime farms by political, bureaucratic and business elites affiliated to the ruling party’ (Hammar and Raftopoulos 2003, 23, my emphasis). In 2012, I spoke to a white former commercial farmer who said:
The under-riding principles of government are to retain control, and political control, over the access of those that they have given land to, whether it is a judge who is expected to give favourable judgements or a simple party cadre expected to roll up and campaign at political times – that's what's happened to the land.
Similarly, in an interview with the BBC, prominent Zimbabwean economist John Robertson said:
They were actually creating a patronage system here. They were saying to people ‘if you display your loyalty to the ruling party, you will get a free piece of land; and now that you've got a free piece of land, if you show any disloyalty to the party – guess what – you'll lose it!
He continued: ‘so this has been a disempowerment exercise – it has been a disempowerment exercise right from the beginning’ (BBC Radio 4, 1 December 2011).
These interpretations of events have not gone unchallenged. In contrast to the ‘mirror images’ narrative, Sam Moyo and colleagues point to numerous continuities in the land occupation movement over time, claiming that it is merely the intensity of land occupations that changed after 2000. Moyo (2001, 314) states that land occupations are part of ‘an ongoing social phenomenon in both urban and rural areas of Zimbabwe, before and after the country's independence … The 2000–2001 occupations mark the climax of a longer, less public and dispersed struggle over land shortages and land demand in the post independence period.’ Sadomba (2008) makes this argument even more strongly by establishing the deep roots of land occupations in Zimbabwe's nationalist and guerrilla movements. Although Sadomba draws distinctions between each of these movements, he treats them as ‘movements within a movement’ because they each shared the same driving force, the desire to liberate the country and reclaim lost lands.
Scoones et al. (2010) also dismiss the suggestion that ‘extensive and multiple’ land grabs have been carried out by political elites, as suggested by Hammar and Raftopoulos. Of course, it is generally understood that the political elite did ‘grab’ some of the best farms and Scoones has been criticised for failing to appreciate the prevalence of this phenomenon outside of Masvingo Province: ‘since the authors’ study sites lie outside the areas where most of the elite cherry-picked their farms, they are unable to examine this proposition. So there is a paucity of data on this point, and the authors confuse the absence of evidence for this proposition as evidence of its absence’ (Kinsey, Unpublished). However, the scale of corruption is often exaggerated. Matondi (2013, 13–14) asks: ‘could 145,775 beneficiaries in the A1 model and 16,386 in the A2 model have all benefited through some form of corruption or nepotism?’ He continues: ‘Such a very large number of friendships would be unprecedented in the history of state development programmes.’
In fact, many early responses to the FTLRP relied too heavily on sporadic media reports, which cannot be relied upon to tell the full story. Empirical investigation was all but impossible save for a select few, such as Marongwe (2003) and Sadomba (2008), who were able to employ certain unique credentials – Sadomba being a war veteran – to access areas where land occupation and redistribution were taking place. The overall lack of empirical evidence has led to the perpetuation of certain misconceptions and generalisations – what Scoones and colleagues problematically term ‘myths’ – which are now being contested (Moyo et al. 2009; Scoones et al. 2010; Hanlon, Manjengwa, and Smart 2013; Matondi 2013). That abuses and violations in the redistribution of land took, and continue to take, place is not disputed; it is the scale of such abuses that is now being contested, as well as the earlier narrow focus on the role of the state and its use of the land issue to hold onto political power. While it is clear that the Zimbabwe African National Union Patriotic Front – ZANU(PF) – hijacked the land issue after February 2000 for its own political ends – as it has repeatedly done since independence, especially during election times – this line of argument has often been too quick to brush aside the idea that post-2000 land occupations also represent a continuation of a century-old popular land movement which has consistently demanded a more socially just distribution of land, albeit with varied degrees of intensity.
So how can we explain the seemingly sudden intensification of land occupations in post-2000 Zimbabwe? In a study of 11 historical cases of peasant mobilisation for land reform – most including some form of land occupations – Huizer (2001) argues that the main precondition for mobilisation was a feeling of ‘extreme frustration’ among the peasantry. This frustration occurred mostly in places where ‘development’ had resulted in the deterioration in the living conditions of the peasantry. More recently, Moyo and Yeros (2005) argue that the prevalence of the neoliberal economic (development) policies has led to the resurgence of rural social movements throughout Asia, Africa and Latin America. They argue that ‘[u]nder the weight of structural adjustment programmes, peasants and workers have seen their conditions of social reproduction deteriorate, giving way to a desperate search for economic and political alternatives’ (1). In Malawi, for example, where the neoliberal economic model dominates, Kanyongolo (2005, 126) writes:
The alienation of peasant customary lands and their conversion to private or state ownership has progressively created and expanded a mass of land-short and landless peasants, who in effect have been presented with the following options: to continue to labour on estates; to migrate to urban centres and become part of the under-class eking out a living at the periphery of the formal market; or to engage in counter-systemic actions, such as resistance against further privatization of communal land and occupations of private or state-owned lands.
Peasant farmers and the unemployed in Zimbabwe were, by the late 1990s, faced with similar options, and it is through this lens that post-2000 land occupations can best be explained and understood.
From liberation to extreme frustration
Zimbabwe attained its independence in 1980 amid high expectations that the new government would introduce land reforms, fulfilling promises made during the war. However, Palmer (1990) notes that ‘the hands of the new Zimbabwean Government were to a large extent tied by the Lancaster House agreement, which was due to last until April 1990’ (166). The constitution ensured that land could only be acquired for resettlement purposes via a ‘willing-seller, willing-buyer’ principle and only under-utilised land could be compulsorily acquired by government. Land made available for resettlement soon became scarce and the market price continued to rise so that ‘where land was offered to Government, in most cases it was expensive, marginal and occurred in pockets around the country, making it difficult to effect a systematic and managed land reform. Moreover, land supply failed to match the demand for land for resettlement’ (Utete 2003, 15). There was also a lack of international support. Palmer (1990, 163) writes that even as the restrictions of Lancaster House were due to expire, ‘there [was] every sign that the British Government [was] striving behind the scenes to perpetuate Lancaster House beyond April 1990 and so prevent significant land reform from taking place in Zimbabwe.’
The priorities of the Zimbabwean government also underwent a transformation in the mid to late 1980s. Dashwood (1996) writes that ‘Zimbabwe's development strategy … evolved from one that gave primary emphasis to equity and meeting the welfare needs of the poor in the early 1980s, to one which evinced a declining commitment on the part of the ruling elite to the welfare of the poor by the end of the decade’ (27). Many have pointed to structural adjustment programmes, but Dashwood's central argument is that ‘[w]hile international influences, in particular, pressure from the international financial institutions, played a greater role in the period 1987–1991, the primary impetus for the shift in Zimbabwe's development strategy was domestic in origin’ (39). It was the ‘embourgeoisement’ of the ruling elite that facilitated the shift in emphasis from social welfarism to economic liberalism. Both Dashwood (1996) and Moyo and Yeros (2005) argue that a new ‘alliance’ was formed between the ruling elite and the – predominantly white – entrepreneurial and agrarian elites, replacing the previous alliance between the ruling elite and the peasantry. Moyo and Yeros (2005, 171), however, assert that the new alliance was ‘cast in the form of “reconciliation”’ but describe it in terms of ‘a reconciliation not “between the races” but “with capital”' (emphasis in the original). The merging of political and economic interests, according to Jenkins, ‘undermined the will to redistribute resources to lower-income groups’ (1997, 594). This echoes Jean-Francois Bayart's (1989) theory of the ‘assimilation of elites’, in which he suggests that powerful groups will form coalitions in order to avoid competition and advance their own respective interests.
The macro-economic situation also deteriorated throughout the first two decades of independence. The introduction of the Economic and Structural Adjustment Programme (ESAP) in 1990–91 had a ‘devastating effect’ on the Zimbabwean economy (Moyo and Yeros 2005, 176), which was both ‘immediate’ and ‘sustained’ (Hammar and Raftopoulos 2003, 6). Deindustrialisation, huge declines in manufacturing, company closures and job losses followed. Interest rates and inflation increased, and real wages declined to below what they were in 1980. The removal of price controls resulted in an increase in the cost of living and per capita income plummeted, leaving most of the population in households unable to meet their basic needs. The poorest cut real expenditure on food, leaving many living below the Food Poverty Line (see Kanji 1995; Hammar and Raftopoulos 2003, 6; Hanlon, Manjengwa, and Smart 2013, 68). Development gains in health and education were reversed: the budget for both was cut and fees were reintroduced for hospitals and schools, resulting in a sharp drop in attendance at both, while at the same time HIV/AIDS was becoming a real problem (Hanlon, Manjengwa, and Smart 2013, 68). To top it all off, the 1991–92 season brought the worst drought of the century.
At the same time, as many were suffering a deterioration in their living conditions as a result of structural adjustment, some did well out of ESAP. As Hanlon, Manjengwa, and Smart (2013, 68) note, ‘[some] company profits increased by 80% in the seven years to 1996, and white farmers did well.’ White farmers did especially well in the export of horticultural products thanks to the devaluation of the Zimbabwean dollar during the 1990s (Ibid.). Those with capital and access to international markets were able to reorient their production from meeting domestic demand to exports, mainly to Europe, and made substantial profits. However, Selby (2006, 194–195) notes that ‘the emergence of young white millionaires against a background of increasing hardships in other sectors implied that the legacies of settler privilege were not abating.’
All of the above combined to produce Huizer's condition of ‘extreme frustration’ among peasant farmers and the unemployed, leading to growing unrest in Zimbabwe. As Hammar and Raftopoulos (2003, 6) correctly observe, the 1990s was ‘a decade of growing public protests, labour strikes and a flourishing of civil society organisations and activities’. A series of alliances was formed between civil society organisations ‘whose collective political voice and effective strike action became increasingly threatening to the regime’ (7). The Movement for Democratic Change emerged from these alliances, representing the first effective opposition political party since the Zimbabwe African People's Union (ZAPU) was absorbed by ZANU to form ZANU(PF) in 1987. The state also encountered open confrontation with the National Liberation War Veterans Association in 1997, when the war veterans demanded a one-off payment and a monthly pension as compensation for their services during the war. They also called for more and faster land redistribution. As Hammar and Raftopoulos (2003, 7) write, an ‘aware[ness] of the veteran's political capital’ and a ‘reluctance to lose their support’ forced President Mugabe to give in to their demands. Although civil society organisations and war veterans were not in alliance with each other, they were at least united in their frustration with the ruling party at the turn of the century, placing ZANU(PF) in an uncomfortable and uncertain position ahead of elections in 2000 and 2002.
The land issue was back at the top of the political agenda. The government organised a donor conference in September 1998 against a backdrop of increasing land occupations. International donors, who saw for themselves the land occupations in Svosve, near Harare, promised financial assistance to re-energise the land reform programme. Land occupations died down in what some might call a ‘wait and see’ attitude. However, the promise of a renewed land reform programme ‘stalled’ because of ‘bureaucratic incapacities and unmet donor conditionalities on the Zimbabwe government side, and perceived hesitations and inconsistencies on the donor side’ (Hammar and Raftopoulos 2003, 9). Frustration reached boiling point and, as Kinsey (1999, 195) prophetically wrote, ‘[t]hose in Zimbabwe most adversely affected by the state's failure to achieve redistribution appear to be increasingly unwilling to sit and listen, before each election, to phrase-mongers who promise resettlement and then fail to deliver. Civil discontent is on the rise.’ The ‘no’ vote in the constitutional referendum might best be described as the proverbial straw that broke the camel's back. If it had been passed, the constitution would have finally untied the government's hands by abolishing Lancaster House, enabling it to compulsorily acquire land for resettlement, paying compensation only for improvements to the land and not the land itself. According to Hanlon, Manjengwa, and Smart (2013, 74), ‘[f]or the war veterans and the landless, the loss at the donors' conference and the defeat of the constitution seemed to imply there would be no land reform.’ Decades of pent-up land hunger finally erupted and veterans began to mobilise people to take matters into their own hands, and – in protest against the state and its failure to meet the promises of independence – they occupied both white-owned and state-owned land.
Motivations of the new smallholders
The preceding argument – that land occupations were the result of extreme frustration among the poor and the landless – is reinforced when we consider the motivations that new smallholders give for joining land occupations or applying for land through official channels (many did both). Also, Scoones et al. (2010, 52) state that knowing what motivated people to take the land is ‘a vital step’ towards knowing who the new smallholders are. During in-depth interviews conducted in 2012, people explained the motivations that drove them to take the land.4 Their responses are summarised in Figure 2.
Reason for occupying/applying for land.
(Source: in-depth interviews, September 2012; N = 166)
It is immediately clear that there was a range of motivations but the most common motivation was to get land. This might seem obvious, but there is more to it than simply getting land. People consistently spoke of ‘more land’, ‘better land’ or land that they could ‘call their own’. In communal areas land is scarce and usually degraded. In the old resettlement areas, the government never made provisions to accommodate the next generation – whether on farms of their own or in formal wage employment – and now, over 30 years later, the children and grandchildren of the original settlers are in need of their own land. In both communal and old resettlement areas, overcrowding has been a persistent problem. In the Zimbabwe Rural Household Dynamics Study (ZRHDS) villages, young men are often found living with their wives and children at their father's homestead, using his field along with the father and brothers. This is quite typical, from a cultural perspective, as women in Zimbabwe often leave their parental home to join the husband's family. When the father passes away, the eldest son usually inherits the plot, leaving other resident sons facing uncertainty. In at least two of the ZRHDS villages, the village head has tried to diffuse the pressure by providing residential stands for married sons within the village boundaries, but these do not come with plots; instead, the sons must continue to share with their fathers and brothers. As young families continue to grow, the situation becomes unsustainable and people are driven to seek alternative livelihoods elsewhere. To recall Kanyongolo's argument, their options become limited and can eventually lead to ‘counter-systemic actions’ such as land occupations.
Mr Kuzanga, for example, heard about the land occupations on the radio. He walked for 100 kilometres from Murehwa, where he was sharing a plot with his father, to Shamva, looking for land. ‘My father-in-law had a big family,’ says Mrs Kuzanga. ‘The plot was small. There were 31 boys. When the land occupations started, my husband went about looking for land … We were told to go to the farms where whites were being sent away’ (interview, September 2012). Similarly, Mr Kapusvu came from Madziva communal area where he and his wife were gold panning because there was very little land for them to farm at home. ‘I came here because I wanted land,’ he said. ‘We were crowded at a small piece of land back in the communal area. We were 18 people on such a small piece of land. We formed a group of people and then decided to look for land’ (interview, September 2012). Mrs Shambare's story also illustrates how the problem can quickly spiral out of control:
We used to stay at a homestead that had 11 boys. My father-in-law had three wives. Out of 11 boys there are 11 daughters-in-law. We are likely to have many boys too. Our father-in-law's plot is small. It wasn't enough for all of us. It won't be enough for our children. (Interview, September 2012)
There were, of course, other reasons for joining the resettlement programme, although none are repeated as often as the desperate need for more and better land. However, that is not to say that the people who speak mainly of their need for land do not also harbour feelings of injustice and resentment toward colonial land expropriation or the failure of government to adequately redistribute land after independence; or that people do not also espouse strong spiritual and ancestral ties to the land. Some of those that got land during Fast Track claim to have fought or collaborated in the liberation war. Mr Johannes, for example, was involved in recruiting young men to fight in the war. He said:
Up to now I don't want to see a white man. He took away my cattle at Batcombe [Farm]. They shared the cattle among themselves. I was shot by a white man at Kafudzaruwa. I went inside … the house was hit by a bullet shot by the Support Unit [and] caught fire. I took a cow hide and covered myself. I managed to escape and run away. One comrade was killed. I did not forget that we needed to take the land. That is why I came here. (Interview, September 2012)
For some, therefore, it was their memories and experiences of the war – or the experiences of others retold – that drove them to ‘take the land’. For others the reasons were religious or spiritual. Mr Mazhoni, an orphan and one of the youngest household heads in the sample at 25 years old, came to the resettlement area in 2005–06 with his grandparents after the spirit of his uncle, who had died fighting in the liberation war, possessed his aunt, who had been mediating the voice of the spirit since 1987, and began to demand land by saying ‘I died for the soil.’ The aunt was taken to the District Administrator and given a vacant plot, which Mr Mazhoni now tends on his own (interview, September 2012).
Indeed, people's reasons and motivations for taking part in land occupations and/or applying for land through official channels are not mutually exclusive, but are multiple, overlapping, perhaps best thought of as situational, in the sense that different justifications are appropriate to different situations. Fontein (2009, 11), for example, writes:
when they [new farmers] first came to occupy land in 2000 … they made no mention of their own autochthonous claims, emphasising only their desire for access to the fertile land available on the farms. It was only later, when other settlers, war veterans and local government officials began to recognise the significance of these ‘traditional’ claims, that they began to announce their own autochthonous claims to the land they were now occupying.
Just as people are constantly changing their political identities to gain access to land (Mkodzongi 2013), they similarly change their justifications for occupying land depending on who they are dealing with and on the prevailing political situation, deploying the reasons that will aid best in legitimising their claims to land. Nevertheless, the driving force was undoubtedly the material need for land as a source of livelihood. In the words of one land beneficiary:
generally, the people felt that they had been short-changed prior to [and post-] independence. They wanted pieces of land and I think that is the bottom line – that people wanted [land], where they could say, ‘This is my own piece of land. This is my piece of land.’ (Interview with Mr Makamure, 2011)
Who are they and where did they come from?
Misconceptions and generalisations surrounding the identities and origins of the new smallholders have persisted over the past decade in large part because of the lack of empirical evidence and the overreliance on sporadic media reports. While it is almost certainly the case that the political elite did ‘grab’ some of the best farms for themselves, this argument is far from revealing of the full story and can often overshadow the fact that the majority of those who received land were ‘ordinary’ people from less well-off backgrounds. In-depth interviews revealed that over half of the households in this sample can be described as ‘ordinary’ people with no clear political connections that would have given them an edge in obtaining land through the FTLRP.
The results presented in Figure 3 are based on responses given by household heads and/or their spouse(s) to questions about their background and employment before coming to the new resettlement areas. The category ‘ordinary’ is difficult to define and its application varies from one study to the next; however, in this case the group were mostly farming small plots in communal areas before coming to the resettlement areas. If one or more of the respondents within a household reported belonging to any other category, then that entire household was automatically excluded from the ‘ordinary’ category. However, one important caveat is that these categories, like people's justifications for occupying land, are not mutually exclusive. Being a war veteran, for example, does not mean that one is not also an ordinary farmer. In fact, many war veterans abandoned that identity and returned to their rural homes after the war – they were not absorbed into the Zimbabwe National Army (ZNA), and remained out of politics at least until 1997. Similarly, civil servants include teachers and health workers who are far from what we might call government ‘elites’ – they have often taken strike action because of the meagre wages they receive. However, there is an assumption that their status might have played some part in their gaining land, so they are placed in separate categories (see Scoones et al. 2010, 53).
As noted, over half of respondents fall into the ‘ordinary’ category. Given that this is by far the largest group, it is worthwhile disaggregating it a little further. If we also take into account area of origin, we can say that just over 70% of ‘ordinary’ farmers come from communal areas, about 25% are from other resettlement areas and less than 5% came from urban areas. Although almost all stated that farming was their main occupation prior to resettlement, some also had low-paid employment as drivers, mechanics, domestic workers, factory workers, general labourers, security guards and gardeners; others engaged in piecemeal work, such as gold panning or casual labour, in order to supplement their income from farming. Nine per cent of respondents can be identified as civil servants, mostly teachers, agricultural extension officers and ministry employees; 7% belong to the security services, mostly serving or having served in the ZRP or ZNA; 6% specifically identify themselves as war veterans, far below the 20% quota often mentioned in the literature; and finally, only 3% identify themselves as former farm workers, a category often overlooked in the redistribution of land. Unfortunately, 20% of respondents were either not available to be interviewed in September 2012, or failed to specify what they were doing before coming to the new resettlement areas. Again, it is worth emphasising that these categories are not set in stone: there may be a few war veterans, civil servants, farm workers or security forces who have fallen into the ‘ordinary’ category. The category of ‘war veteran’ can be problematic, especially given that many war veterans have since abandoned that identity, as noted above.
Although it is difficult to draw comparisons with other studies owing to different applications of the above categories, it is possible at least to highlight some of the apparent similarities and differences. Scoones et al. (2010, 53), for example, found that the majority (almost 70%)5 of respondents in their A1 villagised schemes could be considered ‘ordinary’, which they define as ‘not members of the other categories, and largely asset and income poor’. Moyo et al. (2009, 170) also reported ‘ordinary’ beneficiaries to be the largest group, although they reported a much lower overall percentage of about 21%. However, almost 32% of Moyo et al.'s cases are missing from their analysis, and when these are discounted the percentage increases to almost one-third. Moyo et al. describe the ‘ordinary’ group as having ‘really nothing peculiar’ about them (172). Although there are some differences between findings, these might be explained by different understandings of ‘ordinary’ or differences in sample size, sampling and questioning techniques, or might reflect regional differences in the allocation of land. Nevertheless, they all show the ‘ordinary’ category to be in the majority.
It is also clear from Figure 4 that the vast majority of people coming into the new resettlement areas came from communal areas. This is closely followed by people from different resettlement areas and people from other villages in the same area. In each case, these are rural areas and, when combined, we can say that some 88% of household heads in the sample came from rural areas (excluding those from commercial farms or those who came from households within the same village). This is consistent with the motivations expressed earlier – that most people were coming from rural areas where suitable arable land is in short supply. It is also worth noting that there is no significant difference between sites – in all cases the majority (between 60 and 69%) came from the communal areas, with the bulk of the others coming from other rural areas. Again, comparisons with other studies reveal similar results. Scoones et al. (2010, 53) find that the 59.9% of households that came from rural areas were ‘almost exclusively from nearby communal areas’. Moyo et al. (2009, 21) find that 65.9% of A1 beneficiaries came from communal areas. However, Moyo et al. find that the second largest group – 19.5% – came from urban areas. Again, this might be reflective of different data collection methods, or more likely regional differences.
Table 1 presents a snapshot of the socio-economic characteristics of the four villages in the sample using variables identified by respondents as the most important indicators of wealth and success among smallholder farmers. They are used here in order to reveal a little more of the character of the new resettlement villages.
| North Star | Chiraramo | Golden Star | Bita | |
|---|---|---|---|---|
| Average age of household head 2012 | 45 | 45 | 50 | 44 |
| Year of arrival | 2001 | 2001 | 2001 | 2001 |
| Educational level of household head | Form 2 | Form 2 | Grade 7 | Form 2 |
| Household size (persons) | 6.8 | 7.2 | 8.2 | 5.0 |
| Average land holding (ha) | 5.6 | 5.0 | 5.0 | 3.7 |
| Average area cultivated (ha) | 4.3 | 4.0 | 4.2 | 2.4 |
| Proportion cultivated with maize (%) | 0.51 | 0.53 | 0.43 | 0.63 |
| Cattle ownership (nos.) | 7.4 | 8.0 | 8.4 | 4.0 |
| Other small livestock (nos.) | 2.8 | 1.6 | 4.5 | 3.1 |
| Maize output 2010–11 (t/ha) | 2.4 | 2.5 | 2.5 | 1.0 |
| Maize sales 2010–11 (t) | 3.3 | 2.6 | 2.2 | 0.3 |
| % owning scotch cart in 2012 | 60 | 77 | 82 | 50 |
| % owning plough in 2012 | 80 | 94 | 97 | 82 |
| Quality of housing | Average/Good | Fair/Average | Average | Average |
| % owning a toilet in 2012 | 87 | 94 | 88 | 79 |
Note. Land holdings, area cropped, maize output and sales refer to land owned and crops grown within each village. In all cases the average is reported, except for ‘year of arrival’, which reports the mode. Source: 2012 survey.
There are similarities from one village to another but there are also some significant differences. First, there is a notable difference in the average educational attainment between households in Golden Star and the other three villages. Household heads in Golden Star are, on average, five years older than the others and are less likely to have benefited from the expansion of education in the early years of independence.6 Second, average household size differs with significant variation between NRII and III. It might be that households in NRIII are less able to sustain large families given the lower productive capacity of the soil in that area, and there is also a tendency for people from these families to migrate to urban areas in search of work. Third, there are variations in the average land holdings and area cultivated. In North Star, average land holdings are inflated because one farmer has 18 hectares, which is less common elsewhere.7 In Bita, the average is deflated because some farmers were given irrigated plots of only 3 hectares.8 However, in most cases, households have the standard 5 hectares. Across all sites, people use, on average, between 65 and 84% of their land. Cultivated area was lowest in Bita at 65% and, although farmers there planted the fewest hectares of maize, they devoted the largest proportion of land to food production in 2011–12, averaging 63%. Lastly, some 44% of households had no cattle when they arrived in the resettlement areas, but by 2012 this had fallen to 15% with almost one quarter of households owning 10 or more cattle. Cattle herds in Bita – NRIII – are about half the size of herds in the other three villages, although households do tend to have relatively higher numbers of small livestock – mostly goats and fowls. This is consistent with other findings. Dekker (2009), for example, has shown that households in NRII tend to have greater cattle wealth than those in NRIII, placing them in a better position to maintain livelihood standards in times of drought or other economic hardship through cattle sales.
In addition to identifying people by their background and area of origin, and looking at indicators of wealth at the village level, it is also possible to use these indicators in order to identify subgroups of farmers according to their relative success in terms of productivity, livestock and asset ownership, and so on. However, it is not possible to plot the trajectories of these households in order to determine which are ‘stepping up’, ‘stepping out’, ‘hanging in’ or ‘dropping out’ as others have done (Dorward 2009; Dorward et al. 2009; Scoones et al. 2010). This requires longitudinal data which are not yet available for these households, but which might form the basis for future research. Instead, we can highlight the characteristics of each group, thus bringing us closer to answering the question of who got land during the FTLRP. Using the cluster analysis method, which measures the distance between each pair of variables and groups cases, or households, together based on their similarities to one another and differences from cases that fall into other groups, I have identified four main groups of farmers (see Figure 5). Once again, the variables used in this analysis are those that were chosen by farmers as indicators of wealth and success during wealth-ranking exercises, some of which appear in Table 1. This analysis, however, includes all farming activities, whether they take place within the village or elsewhere.
We know from the ZRHDS data that non-poor households tend to be younger and better educated, are smaller in size, cultivate more land, grow less maize and more cotton and possess more cattle wealth (Kinsey 2010, 19). In this sample of A1 households, we find that ‘Group 1’ is a small group of relatively successful farmers, who, despite being the oldest and least educated with the largest households, were the most productive group in 2010–11, averaging 4.8 tonnes of maize per hectare and selling an average of 8.87 tonnes per household. Note that this was a drought year, and yet everyone in the group sold something. In the 2011–12 season, they had access to and cultivated the most land – some even cultivated more than 15 hectares – and they owned the largest cattle and small livestock herds, averaging 15 and 6.95, respectively. They also appear to emphasise cash crops over food production given that they had the lowest proportion of land cultivated to maize. Perhaps their years of farming experience have compensated for their relative lack of education.
‘Group 2’, on the other hand, is the most educated and the second oldest. They also have the second largest households. In 2010–11, they were joint-second in terms of productivity, averaging 1.96 tonnes of maize per hectare and selling 2.6 tonnes per household; however, unlike Group 1, some sold nothing at all. In 2011–12, they cultivated the second largest land area at 3.76 hectares, half of which was planted to maize. They have the second largest cattle herds but fewest small livestock, averaging 7.33 and 1.91, respectively. They are a middle group of farmers who appear to be doing better than most, but they have not been as successful as Group 1.
‘Group 3’ is the youngest and perhaps least experienced group. They have a lot in common with Group 2, especially in terms of educational attainment, household size, land holdings and productivity; but they have smaller cattle herds and sold the least maize in 2010–11 – in fact, many sold nothing at all. In 2011–12, they also planted the largest proportion of land to maize, indicating that they emphasise food production over cash crops. These might best be described as young opportunists looking to start a new life in the resettlement areas.
‘Group 4’ is distinguishable from all other groups in that it is composed mostly of female-headed households, most of whom are widows. They have the smallest households and the second lowest mean level of education. They have the smallest cattle herds but the second highest numbers of small livestock, averaging 5.17 and 2.38, respectively. In 2010–11, they were the least productive in terms of maize output at 1.43 tonnes per hectare and averaged sales of 1.57 tonnes per household, only marginally more than Group 3. In 2011–12, the group had access to almost exactly the same area of land as Group 1, but cultivated the least of the four groups, with just over half of that area planted to maize. This group appears to be the poorest and although, as a group, they produce enough to feed themselves, some produce nothing at all.
There is, of course, variation within these groups as well as between them, and there are variations between sites too which are not fully captured here. However, I have clearly demonstrated that there exists a spectrum of farmers with different assets and capabilities. The new farmers are not all cronies, nor are they a homogenous group. However, the majority are ordinary farmers who came from the communal areas and now seek to make a living from the land that they received through the FTLRP.
Concluding remarks
The paper has challenged the popular notion that land occupations and the FTLRP, which took place in Zimbabwe after 2000, were merely a state-led campaign against the opposition, or an ‘elite’ land grab. Although ZANU(PF) did use the land issue for its own political gain – as it has repeatedly done since independence – and despite the fact that political elites did grab some of the best farms, post-2000 land occupations also represent a continuation of a century-old popular land reform movement and the vast majority of beneficiaries were ordinary people from communal areas.
This article has shown that a number of factors – the slow pace of land reform, the transformation of government priorities, the deteriorating macro-economic situation, the failure of the donors' conference and the defeated constitutional referendum – combined to produce conditions of extreme frustration by the turn of the century, which ultimately led people to occupy land.
Using survey data and in-depth interviews with 166 newly resettled households, I have shown that most of these new smallholders are ordinary people with no clear political connections, who came from communal and other rural areas to occupy or apply for land because they were no longer able to sustain themselves or their families on degraded and overcrowded land in their former home areas. I have also illustrated the variation that exists between and within new villages. Although most come from less well-off backgrounds, some do better than others; understanding ‘why’ will form the basis of future research.