Changing agrarian relations after redistributive land reform in Zimbabwe

Abstract

Redistributive land reform and agrarian reforms since 2000 progressively changed some of Zimbabwe's agrarian relations, particularly by broadening the producer and consumption base. However they fuelled new inequities in access to land and farm input and output markets. These complex structural changes are explored using a series of surveys, secondary sources and official documents. Findings show that exploitative agrarian labour practices continue despite the diversification of labour towards numerous farms and other enterprises. Agricultural output declined primarily due to reduced inputs and credit supplies, and frequent droughts, but has been rising since 2006. Increasing export production now involves more producers, driven by the diversification of agrarian merchants and contract farming. Agro-industrial capital has gradually increased its domestic operations in the supply of inputs and marketing, especially after re-liberalisation in 2008. Many new farmers accumulate assets although some struggle for social reproduction. Agrarian politics now entail new struggles over agrarian markets, land and labour rights.

Keywords:

• agrarian relations
• agricultural production
• inputs and markets
• farmer organisation

Introduction

Much of the literature on both sides of a polarised debate on agrarian change in Zimbabwe has focused narrowly on the Fast Track Land Redistribution process and its immediate consequences for agricultural production. Many only saw a linear decline in agricultural outputs and loss of formal employment (Robertson 2011 is a typical, recent example of one side on this issue). It is commonly claimed that Zimbabwe had been the breadbasket of Southern Africa. Yet it was an irregular food exporter and importer while South Africa met regional food deficits (Moyo 2010). The decline in output is attributed, often implicitly, merely to the replacement of skilled large-scale farmers with subsistence producers and the loss of private property rights (e.g. Richardson 2005). This assumption tends to point to an alternative presently being proposed to ‘rationalise’ FTLR structures through green revolution-type reforms which integrate small farmers into agribusiness (USAID 2010).

These perspectives neglect to examine broader issues of the changing uses of land and labour by varied farming classes with differentiated access to agrarian markets. The displacement of white farmers and farm workers is highlighted (Hammar et al. 2010) but changes in the agrarian labour relations are not (see Chambati 2011 in this collection). Moreover, changing agrarian relations in a labour reserve economy are inevitably not only a consequence of what happens in relation to land reform but also of what is happening with urban production and labour markets. There are few efforts to examine changes in the agrarian markets (e.g. as does Scoones et al. 2010) and the recent strategies of capital in response to shifts in policy and global markets (Moyo 2011b). In addition to the issues of changing agrarian relations and the impact of agricultural markets and capital, there is a need to study state intervention carefully to elucidate its contribution to agrarian change. Instead media-based reports tend to be restricted to highlighting corruption and patronage in favour of ZANU-PF (Zimbabwe African National Union-Patriotic Front) elites and the exclusion of opponents, missing the substantive class and regional dynamics in which these are embedded.

The Government of Zimbabwe argues that it supported farmers and intermediaries in a non-partisan manner, in the face of droughts, structural constraints on credit and fiscal capacity and political isolation and sanctions from Western nations (RBZ 2007a). Such constraints existed but policy implementation was riddled by incoherence, inconsistencies and class contradictions (Moyo and Yeros 2007, 2009). The underlying issue was how to finance the strategy in favour of peasants while retaining state autonomy.

These questions require empirical attention in the context of a progressive vision of transforming settler-colonial agrarian relations. Land redistribution alone does not automatically yield egalitarian agrarian relations. Progressive agrarian reforms seek increased productivity among small producers to increase food and other supplies to home markets and contribute to industrial diversification and increased employment (see Patnaik 2003). Articulated national development requires trade protection and subsidies (Chang 2009). Implementing such a vision requires producer cooperation (Moyo and Yeros 2005) against dominant capital, which prioritises externally-oriented production and markets while depressing commodity prices.

Articulating such a perspective, this paper provides a macro-picture of Zimbabwe's changing agrarian relations since 2000. After outlining the agrarian history and policy context (detailed in Moyo and Yeros 2005 and Moyo 2011b), it describes the unfolding land and labour relations (also detailed in Moyo 2011a and Chambati 2011 in this collection) to illuminate the agrarian class structure. The orientation of agricultural production is then examined to elucidate its class and regional character. This enables us to explore the trajectory of uneven access to agrarian input and output markets. This highlights the reconfiguration of agrarian markets and the re-insertion of capital through contracts focused on export-oriented production and the class differentiation of farm investment and productivity. Finally, the paper looks forward and projects possible future outcomes by outlining the way farmers are reorganising themselves for state support, and access to markets and land (see also Murisa 2011 in this collection).

This paper relies on research undertaken through the African Institute for Agrarian Studies (AIAS) between 2002 and 2010 on land reform beneficiaries and wider agrarian processes. Various district surveys were undertaken between 2002 and 2004,1 and a national baseline survey was administered in six districts between 2005 and 2006.2 This captured data on land tenure, socio-economic characteristics of beneficiaries, production, markets, labour and farmer organisational issues from 2089 land beneficiaries and 761 farm workers. That data is cited as AIAS (2007), while its analysis was reported in Moyo et al. (2009). Follow-up field work undertaken between 2007 and 2009 is also cited.

The context of agrarian reform

Equitable agrarian reform in Zimbabwe was compromised during independence negotiations in 1979 in favour of power transfer and liberal democratic reform (Habib 2011). Settler-colonial accumulation by dispossession from 1890 created a labour reserve economy (Amin 1972) dependent on cheap domestic and foreign migrant labour (Arrighi 1973). Peasant farming, rural small-scale industry and commerce were repressed through extra-economic regulations and taxes, but this did not create full-scale proletarianisation (Bush and Cliffe 1984, Yeros 2002). Racial and class inequalities in the agrarian relations were consolidated by discriminatory subsidies to large-scale farmers (Moyo 2002) and narrow import substitution export-led strategies. The consequent rise and fall of the peasantry is well documented (Weiner 1988).

Independent Zimbabwe inherited a racially skewed agrarian structure and discriminatory land tenures dominated by 6000 white farmers and a few foreign and nationally owned agro-industrial estates, alongside 700,000 peasant families and 8000 small-scale black commercial farmers (Table 1). From 1980 Zimbabwe pursued a market-based land reform programme. By 1990 land redistribution had resettled over 70,000 families on about three million hectares (Cliffe 2000). As far as this went it was considered successful in meeting production and livelihood targets (Cusworth and Walker 1988) but the land redistribution was inadequate (Moyo 1995). It left regressive agrarian labour relations on the 75percent of large farms which were not redistributed (Amanor-Wilks 1995).

Table 1. Agrarian structure: estimated landholdings from 1980 to 2010.

	Farms/households (000s)						Area held (000 ha)						Average farm size (ha)*
Farm categories	1980		2000		2010		1980*		2000*		2010*		1980	2000	2010
	No	%	No	%	No	%	Ha	%	ha	%	ha	%
Peasantry	700	98	1,125	99	1,321	98	16,400	49	20,067	61	25,826	79	23	18	20
Mid-sized farms	8.5	1	8.5	1	30.9	2	1,400	4	1,400	4	4,400	13	165	165	142
Large farms	5.4	1	4.956	0.4	1.371	0.1	13,000	39	8,691.6	27	1,156.9	4	2,407	1,754	844
Agro-Estates	0.296	0.1	0.296	0.02	0.247	0.02	2,567	8	2,567	8	1,494.6	5	8,672	8,672	6,051
Total	714	100	1,139	100	1,353	100	33,367	100	32,726	100	32,878	100	46.7	28.7	24.3
Source: Adapted from Moyo 2011a (table 1). Notes: ^*The average farm size of the peasantry includes common grazing lands.

The adoption of the Economic Structural Adjustment Programme (ESAP) in 1990 further slowed down land redistribution, encouraged renewed land concentration and foreign land ownership and fuelled export-oriented production, including extensive landholdings for eco-tourism (Moyo 2000). State agrarian subsidies and social welfare transfers to peasants were reduced. This reversed the production gains realised by the top 20percent of the peasantry, leading to declining maize yields from 1991 (Andersson 2007). ESAP exposed farmers to volatile and monopolistic world markets and reinforced unequal production relations (Moyo 2000).

Wider social dislocations emerged as fiscal capacity dwindled (Bochwey et al. 1998). Increased rural landlessness and retrenchment of urban workers extended land hunger (Yeros 2002). Wage repression led to extensive strikes and protests by industrial and agricultural workers between 1994 and 1998 (Sachikonye 2003, Rutherford 2003). Unprecedented political conflicts emerged within and outside the ruling ZANU-PF (Moyo and Yeros 2007). Authoritarian rule escalated and elections, involving the newly formed MDC, were bitterly contested from 2000, often with violence (Raftopoulos and Mlambo 2009).

These contradictions ignited popular land occupations from 1997 (Moyo 2001), many of which were led by liberation war veterans (Sadomba 2008). Many of these entailed violence and most white farmers were replaced (JAG/RAU 2008). Government condoned the occupations but sought to co-opt them (Moyo 2001). Eventually it gained control over them through state land expropriations and by redistributing land to over 150,000 families in two types of schemes under the Fast Track Land Reform Programme (GoZ 2001).3

This departure from neoliberal prescriptions on land reform only partially contested the prevailing wisdom that agricultural growth and viability require large-scale farm sizes (Cousins and Scoones 2010) since the redistribution programme provided for black commercial farmers on relatively large plots, although these are much smaller than those of the erstwhile white farmers. The reliance on cheap labour was not abandoned as A2 farms presumed labour supplies from landless workers. Nonetheless, the process rowed against the current of escalating land alienation in Africa and drew political opposition and Western sanctions.

Facing external pressures and economic decline, the Zimbabwe state radicalised economic policy between 2001 and 2007 while negotiating ‘normalisation’ with capital by allowing agro-industrial estates and agri-businesses to operate within a heterodox policy framework (Moyo and Yeros 2007, Moyo 2011b). State controls over agricultural commodity markets, trade and financial markets were introduced in 2001 (World Bank 2006). Subsidies targeted distressed industries, including agro-industries to improve the supply of inputs to farmers (RBZ 2006). Agricultural inputs and food prices were regulated. The parastatal Grain Marketing Board (GMB) monopolised grain buying.4 Genetically modified seeds were actively prohibited and open-pollinated seed was encouraged. Mechanisation subsidies to counter labour shortages and expand cropped areas were escalated from 2006. The state attempted to expand agricultural production on public estates (Moyo 2011b). The Look East Policy was introduced in 2005 to diversify foreign markets. Agrarian labour policy however allowed for low wages (Chambati 2011).

The increasing shortages of foreign currency for imports and of consumption goods – the result of declining production and rising interest rates, which made credit unaffordable – elicited expansionary fiscal interventions. The excessive printing of money and the uncontrolled use of parallel currency markets by the state and businesses generated hyperinflation. This in turn led to aggressive price controls and fuelled underground markets, while sanctions were escalated.

These simultaneously defensive and proactive policies did not extensively oust capital nor fully socialise production relations partly because of the embeddedness of some ruling party leaders and other influential actors in capital and the rootedness of commodity production relations. The plan was opposed ideologically and in practice by big business (see Moyo 2011b). Due to conflicting class and political interests, factions of the ruling party elite and other factions of domestic capital and international capital (represented by black nationals) clashed over the allocation of widely spread and limited public resources and subsidies. These measures affected the interests of opposing classes and politicians, with some seeking to evade or benefit from them. Planning deficiencies were evident. Corruption emerged within and outside ZANU-PF as various classes competed for access to the subsidies and rents. Patronage often included or excluded both political opponents and supporters. These events widened the fractures within ZANU-PF and fuelled the violently contested elections of 2008 (Moyo and Yeros 2009).

Facing dramatic economic collapse and a political stalemate the economy was liberalised in 2008 (RBZ 2009) and a ‘power-sharing’ government was formed through regional mediation (GPA 2008). Controls on agricultural markets, the capital account and trade, and off-budget subsidies were abandoned, and the economy was ‘dollarised’ (RBZ 2009). Foreign investors were invited, albeit within the ‘Indigenisation’ policy requiring domestic control of majority shares (GoZ 2011). These policy shifts, which resulted from the agency of diverse landholders and classes, subsequently elicited diverse responses from them. This complex, contradiction-riddled and confrontational process reshaped various dimensions of the agrarian relations which had been fundamentally restructured by land redistribution.

Changing agrarian structure: land and labour relations

Fast track land redistribution undermined the underlying logic of settler-colonial agrarian relations founded on racial monopoly control over land that deprived peasants of land-based social reproduction and compelled cheap agrarian labour supplies. Redistribution reversed racial patterns of land ownership and broadened access to land across the ethnically diverse provinces, while replacing most private agricultural property rights with land user rights on public property (Moyo 2011a). This represents socially progressive agrarian change. But it also generated new inequities based on uneven land ownership and control of labour.

Zimbabwe's agrarian structure comprises four relatively distinct farm categories, based on differences in land size, forms of land tenure, social status of landholders and capacity to hire labour. These include peasants, also called small producers here; middle-sized farmers; large-scale capitalist farmers; and agro-industrial estates, plantations, and conservancies. Redistribution enlarged the peasantry and expanded the number of mid-sized farms, while downsizing the number, farm size and area of large-scale capitalist farms, as well as of the agro-industrial estates (Table 1). Landlessness remains, especially among some farm labourers. This structure has created a platform for new agrarian class formation processes and struggles.

Agrarian relations among the peasantry continue to be defined mainly by self-employment of family labour towards producing foods for auto-consumption and selling some surpluses, as well as various non-farm work and short-term wage labour. The peasantry have differentiated capacities to hire limited labour, and some provide labour services to others. Most of the families hold customary rights to arable and homestead plots and common grazing areas in Communal Areas, while their A1 beneficiary counterparts hold state permits for similar family and common land rights. About 30percent of the A1 beneficiaries are made up of urban workers and a few former farm workers (Moyo et al. 2009). Eighteenpercent of the land beneficiaries retain homes and plots in Communal Areas to diversify their reproduction and production; using extended family resources and sharing land with extended family members and subletting to others is commonly practised (Moyo et al. 2009).

Over 31,000 middle-scale and large-scale capitalist farmers, most of whom are blacks, now exist. Two thirds of them received land as A2 beneficiaries in all provinces on varied land sizes (Table 2), with varied farm assets (Table 5). These farmers rely on relatively larger amounts of hired labour than on family labour (see Chambati 2011). They hold land through tenures amenable to market transactions, mainly through leases, while a few retain freehold title. The majority of them originate from the middle class, including currently or formerly employed professionals, small non-farm capitalists and rural ‘elites’, including chiefs and some better-off peasants, as well as some working class people (AIAS 2007). Those with larger-scale farms tend to be better educated and linked to employment and business (Moyo 2011a), and are better placed to negotiate political power and mobilise resources. A few hire farm managers, while some rent land claiming their land sizes are too small to be ‘viable’ (AIAS 2007). Some hold multiple farms (Moyo 2011a).

Table 2. National classification of (A2) farms.

	Farm sizes
Provinces	below 100	101–300	301–500	501–1500	1501–3000	3001–5001+	Total
Manicaland	1,427	404	100	31	2	1	1,965
Mashonaland Central	1,991	696	184	60	5	3	2,939
Mashonaland East	3,113	1,049	195	95	4	3	4,459
Mashonaland West	6,353	1,421	300	158	16	8	8,256
Masvingo	1,027	86	43	121	49	4	1,330
Mat North	198	221	66	145	35	8	673
Mat South	117	643	321	230	54	1	1,366
Midlands	136	606	428	191	15	9	1,385
Total	14,362	5,126	1637	1,031	180	37	22,373
Source: Compiled by author from MLRR (2010) data.

The agro-industrial estates were reduced to 240 establishments, mostly owned by large-scale capital and covering over one million hectares or three percent of all the farming land (Moyo 2011b). They still hold freehold title to land in vertically integrated enclaves, including tourism conservancies and state estates. They hire large amounts of permanent and seasonal labour (Chambati 2011) and contract an expanding number of outgrowers. The latter comprise small and medium-sized land beneficiaries relying on family and hired labour. The state has retained its plantations and expanded production through partnerships with capital. The indigenisation policy calls for the redistribution of the estates' shareholdings to locals. So far about 39 blacks hold shares in conservancies (The Herald 2011a).

The persistence of large-scale landholdings meant the exclusion of potential land reform beneficiaries (Moyo 2011a) and fuels the ‘illegal’ occupation of lands (MLLR 2009). The policy of limiting access to redistributed land for former farm workers was partially motivated by the desire for cheap labour supplies. Landless people and poorer peasants still provide some farm labour services at low wage rates (see Chambati 2011 in this collection). Many landless farm labourers reside precariously on new landholdings, perpetuating exploitation practises via tenancy. However, once the pre-2000 relations of agrarian labour were undermined by land redistribution agrarian labour shortages became more common on the remaining large-scale farms. The number of full-time labourers on capitalist farms has declined and short-term hired labour has expanded on diverse farms and in rural non-farm activities (Chambati 2011, this collection). About 25percent of the beneficiaries in our surveys retained other jobs, while 27percent of the land beneficiary household heads are non-residents who combine urban waged labour with farming (Moyo et al. 2009). Policies intended to curb widespread petty natural resource and mineral exploitation by the landholders, landless and others by promoting farm mechanisation failed to address such labour shortages (Utete 2003). These trends confirm that repeasantisation and semi-proletarianisation are simultaneous outcomes of the agrarian reforms (Moyo and Yeros 2005) although we cannot foreclose the trajectory (see O'Laughlin 2002).

Unequal land and labour relations are being consolidated by tendencies towards land concentration through informal land rentals. About 25percent of the land beneficiaries sublet or share their land (Moyo et al. 2009) without official sanction. Some of these households lack production inputs or face social calamities such as illness or deaths (Mhondoro field interviews 2008). Others sublet land for speculative reasons or seek to maximise incomes from farming partnerships (Mhondoro field interviews 2008). These tendencies suggest the contingency of such land markets. Large-scale re-concentration of agricultural lands is however restricted by state ownership of redistributed lands and natural resources. Moreover A2 farmers have no legal right to evict informally settled farm workers.

Various A2 farmers advocate the commodification of land through freehold tenure, land rentals and sometimes by evicting land beneficiaries but this is actively resisted by small landholders. However, 80percent of the surveyed beneficiaries perceived their land tenure to be secure. About 16percent of them had been threatened with eviction and many had successfully resisted this, while seven percent had once been evicted, mostly from A1 and highly prized peri-urban landholdings (Moyo 2011a). However, land conflicts mainly arise from unclear boundary definition, double allocation of some plots by the state and disputes over the use of common natural resources. This affected 21 percent of the beneficiaries in 2006, mostly in higher agro-ecological potential and peri-urban areas, some of whom putatively retained Communal Area homes as insurance against eviction (Moyo 2011a).

Redistributive land reform did not reverse the regressive agrarian relations evoked by patriarchal power relations. Land access biases against women, youth and immigrants and the exploitation of female labour through male control of products are common (see also Makura-Paradza 2010). While more women secured their own land than in previous reforms, men and husbands still dominate agrarian transactions (WLZ 2007, Moyo 2011a). The extension of traditional leadership to newly redistributed areas widens these contradictions (Murisa 2010), although this cannot be over generalised. Legally, however, the new land user rights are derived from the state and not custom. Agrarian relations are still coloured by power relations derived from ethno-regional identity. Land redistribution re-linked people with their original ‘homes and ancestral spirits’, providing scope to re-mobilise lineage-based ethnic ties and territoriality (Mazoe focus group discussion 2005, Mkodzongi 2011). Often these affinities are used to exclude those defined as not belonging, but this operates unevenly among the provinces and peri-urban areas (Moyo 2011a).

Despite the progressive outcomes of land redistribution, agrarian relations are imbued with salient struggles over access to land and labour. These are amplified in various production processes.

Changing agrarian production relations

Extensive land redistribution is expected to alter the structure and orientation of agricultural production, including generating a transitional decline in output as new classes of farmers mobilise resources to establish themselves in relation to changing markets and state interventions. Indeed, the output of Zimbabwe's main agricultural commodities started declining in 2002 and then selectively began to rise from 2006. The number of farmers involved in producing diverse commodities and the overall cropped area expanded substantially, while yields generally declined. This trend affected the commodities differentially as some had been predominantly grown by small-scale farmers, while others had been grown mainly on large-scale farms and plantations on land which was unevenly redistributed. Moreover some commodities are produced predominantly for export markets and are externally financed. A new uneven class and regional production structure has emerged restructuring some of Zimbabwe's social relations of agrarian production, while consolidating others.

Expansion of food production among the peasantry

The output of maize, Zimbabwe's main staple grain produced mostly by peasants, declined severely in an erratic long run pattern associated with droughts (Figure 1). National maize yields per hectare fell to about 50percent of the 1990s average, while the average yields of A1 maize producers were half those realised by A2 producers, and land beneficiaries in wetter agro-potential areas realised twice the yields of those in drier regions (AIAS 2007). However, the output of small food grains such as sorghum and millet increased (Figure 2) although their average yields fell 40percent below the 1990s average (MIMAD 2010a). The output of wheat, which before 2000 was predominantly grown by large-scale farmers, declined dramatically (Figure 2) and the area cropped to wheat fell from an average of 58,000 hectares in the 1990s to 18,200 hectares in 2010.

Figure 1. Sub-sectoral maize production trends (1980–2010) in Zimbabwe. Source: MAMID (2010a, 2010b).

Figure 2. Cereal crops output: 1990s vs. 2000s average. Source: FAO (2009), MIMAD (2010a, 2010b, 2011).

However, the numbers of food grain producers expanded after the Fast Track Land Redistribution Programme, leading to a major increase in the national cropped area dedicated to food grains from 1,794,527 hectares in 1999 to 2,655,687 hectares by 2011 (MAMID 2010a). Land reform beneficiaries dedicated 78 percent of their cropped land to food grains (Moyo et al. 2009). This shifted the orientation of production and use of prime lands away from exports to the staple grains prioritised by peasants.

The outputs and cropped areas of oilseeds such as soyabeans and sunflowers, which target the home market but were mainly produced by large-scale farmers before 2000, also declined but later they experienced a limited up-turn (Figure 3). The output of groundnuts and edible beans, however, increased, and these crops continued to be grown mainly by peasants. About 21percent of the land beneficiaries grew groundnuts, while around sixpercent of them grew beans and soyabeans (Moyo et al. 2009). This suggests that the numbers of higher value food producers has expanded and that peasant production is diversifying.

Figure 3. Oilseed output: 1990s vs. 2000s average. Source: TIMB Statistical data, Cotton Ginners Association (2010), MAMID (2011).

Resilience of export oriented production?

In terms of financial value, the largest share of agricultural output decline occurred among export commodities which had been predominantly produced on large-scale landholdings. In 2004 tobacco output had fallen by 72 percent from the 1990s average (Figure 4) but by 2011 its output was rising substantially (Figure 4). Its yields and planted area declined by 55 and 20percent respectively during the period and fertiliser utilisation per hectare was halved by 2004 (TIMB2010,MIMAD2011). But over 50,000 farmers, including 30,000 peasants, are now producing tobacco on smaller cropped areas, compared to about 700 large-scale and 3,500 small producers around 2002 (TIMB 2010, MIMAD 2011). The A2 farmers are realising higher yields as they use more inputs (Moyo et al. 2009). Some remaining white farmers continue to produce tobacco.

Figure 4. Key exports' output trends: 1990s average vs. 2000s.Source: TIMB Statistical data, Cotton Ginners Association (2010), MAMID (2011).

The output of cotton, which is largely exported and was predominantly produced by small farmers, surpassed the 1990s average by 49percent by 2011 (Figure 4). This reflects a sizeable expansion of small-scale producers in a commodity whose production was already entrenched, the cotton's tolerance of drought, and the continuity of contract-based inputs supplied by capital. On average fourpercent of the land beneficiaries grew cotton, while 21 percent grew it in Chiredzi district, despite its limited cotton growing history (Moyo et al. 2009). Established agro-industrial production structures and technocratic wisdom were being challenged by farmers in the new milieu (see also Scoones et al. 2010).

The outputs of plantation export commodities only began to decelerate around 2004 but these were rising by 2011 (Figure 5). Sugar output fell by 20percent from the 1990s average levels in 2006 and then by 50 percent during hyperinflationary conditions between 2007 and 2008, only for the rate of decline to decelerate by 2011 (EU 2009, RBZ 2011). The structure of sugar production hardly changed as the area cropped by the estates was hardly reduced while the outgrowers' cropped area declined substantially (EU 2009, RBZ 2011). The number of sugar outgrowers grew from 47 whites with an average of 147 hectares each in 2000 to about 560 with average hectarages of 10 to 30 hectares (Moyo 2011b).

Figure 5. Tea and coffee output trends in Zimbabwe. Source: IMF (2005), Zimbabwe Tea Growers Association data, Zimbabwe Coffee Mills data.

The number and area of large-scale farm and estate producers of coffee and tea had declined by 2006, but tea output patterns were anomalous compared to other crops. By 2010 tea output levels were well above the 1990s averages, but this level was lower than the peak reached prior to 2007 (Figure5). Meanwhile coffee output had mainly declined by over 90 percent in 2010, with its cropped area falling by over 30 percent, among the outgrowers, whose numbers almost doubled (Zimbabwe Coffee Mills).5 Remaining tea estates are now diversifying production towards macadamia nuts, pineapples and passion fruit, ostensibly due to labour shortages and lower prices, while outgrowers in tea and sugar growing areas are also producing some foods due to input shortages (EU 2009, USAID 2010).

The production of formally marketed dairy and beef was previously dominated by large-scale farms. The national cattle herd fell by 19percent by 2009 (Table 3). Most cattle are now held by many smaller-scale and middle-scale farmers practising mixed farming, which provides various foods, draught power and manure. Yet 58percent of the land beneficiaries had no cattle, while 36percent had between three and above 20 cattle, and the rest had one to two beasts (AIAS 2007). However, 70percent of the households in drier agro-ecological regions, officially considered ranching areas, had cattle. Milk output also fell by 66percent from 2001 to 2010 (MAMID 2011).

Table 3. Cattle numbers by farming sector: 2001 – 2011.

Sector	2001	2003	2005	2009	2011
A2/LSCF	1,291,110	453,418	519,028	442,080	453,385
Communal	4,398,081	3,994,830	3,604,361	3,692,196	3,529,739
Resettlement & A1	505,360	717,969	844,800	919,616	1,020,070
Small scale	23,565	180,648	219,424	167,828	147,559
Total	6,418,116	5,296,865	5,187,613	5,221,720	5,156,753
Source: MAMID (2010a, 2010b), FAO (2009).

There were rapid declines in pork production by 2005 but pork output had risen substantially by 2009, while the numbers of goats were stable compared to sheep, which are not a common food ( Figure 6). Less than 25percent of the land beneficiaries had small ruminant livestock, while twopercent of them had piggeries by 2006 (MAMID 2011). Formally marketed pork production was previously dominated by over 100 LSCF producers and agro-industrial plants, but by 2010 it involved over 250 smaller producers.6 Overall the diversification of livestock producers was accompanied by lower quality, breeding stocks and calving rates since investments in breeding, animal health and pen-feeding had declined (FAO 2009).

Figure 6. Small stock production. Source: MIMAD (2010b).

Diversification of the production base and class bias

In general agricultural production patterns during the 2000s became more differentiated in class and regional terms, while export-oriented output rose faster than food. The production base was restructured by introducing more producers into all commodities and expanding the overall cropped area substantially, despite the decline of yields for most crops and livestock. Numerous producers earned farming incomes and provided their own food. A process of income re-distribution was underway, although this favoured an expanded range of middle- to larger-scale farmers, mainly in wetter regions. While more farmers are now producing exports, they are however well below 15 percent of all the farmers.

Average land utilisation rates among land beneficiaries were at 40percent (AIAS 2007) comparing favourably with former large-scale farming areas (World Bank 1991). Around 2006 about 54percent of the beneficiaries cropped less than three hectares, while only 14 percent cropped more than 10 hectares (Table 4). Thus middle-scale and larger-scale landholders cropped proportionately less land than the peasants. Their pre-2000 counterparts usually cropped below 700,000 hectares, despite employing more formal labourers (World Bank 1991).

While before 2000 agricultural production was predominantly export oriented, the incomes realised were concentrated among a few large farmers, alongside domestic and foreign capital. Peasants cropped much more land and used more labour towards producing various foods largely for auto-consumption. The latter trajectory has been consolidated, but land and labour productivity continue to be low. Consequently, domestic food production has on average been 35percent short of requirements, with wheat, soyabeans and dairy faring worst, while the output of pulses has expanded. Zimbabwe continues to have some relatively more secure food enclaves, reflecting uneven production and productivity patterns. Small producers in the southern districts continue to face regular grain deficits, while those in the wetter regions increased their production of pulses and cash crops, particularly tobacco and various vegetables.

Class and regional biases in the capacities of various farmers to produce relatively larger areas of high value crops and export commodities are being reproduced. A broader base of capitalist farmers has emerged, while the plantations are consolidating their vertical integration into world markets using more outgrowers. More peasants and middle-scale producers are now slowly expanding the supply of more diverse foods and raw materials to the home and export markets. This also entails the reinsertion of large foreign capital into Zimbabwe's restructured agrarian markets.

Differentiated access to agricultural inputs and markets

Agricultural productivity generally declined due to reduced and uneven access to inputs and output markets. This particularly affected smaller producers who nonetheless deployed their labour to expand cropped areas. Access to inputs was also constrained by reduced public and private agricultural finance, leading to the diversification of input supply and commodity marketing arrangements by capital.

Uneven access to agricultural inputs

Maize seed was in short supply during 2003 and 2006. Three transnational companies had dominated hybrid maize seed production by 1999 through contracts with about 200 larger-scale growers whose land was redistributed. By 2010 numerous medium- and large-scale farmers were being contracted to produce seed, unravelling the previous oligopoly. Similarly many more middle-scale farmers were producing tobacco seedlings and meeting current demand (TIMB 2010). Shortages of potatoes and vegetable seeds persisted, leading to supplementary imports. By 2011 over 80percent of all the farmers were using commercial hybrid seeds now grown by more farmers contracted to capital, which retained dominance in the privatised bio-genetic industry.

Fertiliser application in Zimbabwe in the 1990s averaged 30 kg/ha. This was halved by 2004 (FAO 2009) and applied to larger cropped areas. Around 2006, 50 percent of the land beneficiaries were utilising inorganic fertilisers, mostly for maize, tobacco and cotton production, with relatively more A2 beneficiaries using fertilisers, and 20percent of A2 beneficiaries using pesticides (Moyo et al. 2009). The wetter agro-ecological regions used more fertilisers. Nationally less fertiliser was applied by small farmers than by the larger farms (Figure 7). Export crops used the largest share (Tripathy et al. 2007). Most A1 farmers were using animal manure rather than fertilisers (Moyo et al. 2009), bearing in mind the rise of inputs prices globally (Moyo 2010). Less than 10percent of farmers also adapted to the rising price of inputs and access problems by adopting conservation farming to optimise absorption of water and fertiliser (FAO 2011).

Figure 7. Sub-sectoral utilisation of locally produced fertiliser. Source: Triparthy et al (2007).

The use of productivity enhancing inputs on livestock was mainly found in the drier southern provinces. Twelve and 21percent of the land beneficiaries used stock feeds and veterinary chemicals respectively, although 34percent of the pork producers used them (Moyo et al. 2009). Only 35percent of the cattle producers used public dipping and veterinary services, with slightly more A1 beneficiaries using public dips (Moyo et al. 2009). This limited livestock productivity. The re-insertion of large capital in the livestock sector was relatively limited because of the continued and dispersed control of breeding stock by former large-scale farmers and the pervasiveness of uneven regional investment.

Agricultural production continues to depend mainly on rain-fed farming. Only five percent of the national cropped lands are irrigated and plantations control 57percent of this, while small producers control 30percent (World Bank 2006). About 17percent of the land beneficiaries had one form of irrigation facilities, while 28percent of the A2 farmers had irrigated crops compared to 14 percent of A1 farmers, and only 10percent of both groups had invested in irrigation (Moyo et al. 2009). Irrigation was slightly more common in Chipinge and Chiredzi (Moyo et al. 2009) where more irrigation facilities were pre-existing on plantations. Some irrigation facilities were disabled by departing landowners and land occupiers and numerous dams remain underutilised (World Bank 2006). This uneven class and regional distribution of irrigation facilities is also associated with export production.

Agricultural productivity is also constrained by low and uneven access to farm machinery. Most peasants still depend on labour-intensive ox-drawn traction and hand weeding. Only 49percent of the land beneficiaries had access to animal-driven ploughs, while less than 20percent had access to power-driven equipment, with only six percent of the A1 beneficiaries having access to tractors, compared to 36percent of the A2 farmers (Moyo et al. 2009). Over 70percent of the A2 farmers who used tractors owned them (ibid). Small landholders own less than 22percent of the national tractor fleet, farm equipment and machinery (MAEMI 2009). Public or private draught power hire services are limited, and the government's mechanisation programme added only 3217 tractors to the national stock (MoF 2010). During the 2000s limited fuel subsidies were provided, but these mainly benefited A2 farmers.

About 30percent of the land beneficiaries had on-farm and off-farm infrastructure, including some left on the redistributed farms (Table 5). Some A1 farmers shared farm houses and stores for social services, while A2 farmers gained these individually. Given the limited availability of credit, these investments are significant. While access to subsidised inputs during the 2000s reached diverse farmers, the outcome favoured larger-scale export farmers with better access to markets (AIAS 2007). Thus, only some resource rich farmers had access to inputs and this limited the capacity of most to hire labour (see Chambati 2011) and invest.

Table 4. Total cropped area by farm size in selected new resettlement areas.

	Farm sizes (ha)
Cropped	1-19		20-49		50-99		100-299		300+		Total
area (ha)	No.	%	No.	%	No.	%	No.	%	No.	%	No.	%
0	177	18.8	122	19.8	30	16.3	49	27.5	11	37.9	389	19.9
0.1-1	90	9.5	44	7.2	13	7.1	27	15.2	–	–	174	8.9
1.1-3	285	30.2	145	23.6	27	14.7	37	20.8	3	10.3	497	25.5
3.1-5	222	23.5	105	17.1	24	13.0	23	12.9	3	10.3	377	19.3
5.1-10	97	10.3	92	15.0	40	21.7	12	6.7	–	–	241	12.4
10+	73	7.7	107	17.4	50	27.2	30	16.9	12	41.4	272	13.9
Total	944	100.0	615	100.0	184	100.0	178	100.0	29	100.0	1950	100.0
Source: AIAS Household Baseline Survey (2007).

The capital intensity of farming is uneven and influences wider agrarian relations. Those few larger-scale farmers using motorised traction include both high and low intensity labour hirers, while the majority hire little labour and are poorly capitalised (see Chambati 2011). Class biases in the control of land, labour and access markets have no doubt also been shaped by unequal political connections and social status, as well as the re-configuration of such markets.

The reconfiguration of agrarian markets

The production of farm inputs such as fertilisers by domestic industry was also falling by 2003, as were imports (Figure 8). Many agro-industries did not adapt to the changing demand structure. Agrarian merchants and fertiliser producers reduced local operations, and increased operations in neighbouring countries. As interest rates rose, agri-business increasingly depended on subsidised foreign currency and credit in exchange for reduced prices and turned to manufacturing inputs on pre-paid contracts for large-scale exporters.

Figure 8. National fertiliser supply (2001 – 2010). Source: Compiled by AIAS from Zimbabwe Fertilizer Company (ZFC) 2010 data.

More fertiliser and tractors were imported from China and Iran through concessional loans and from South Africa and elsewhere in cash. New indigenous importers were contracted to supply government input schemes. Similarly, domestic producers and suppliers of machinery and implements increased supplies when contracted by the government (see RBZ 2007a). This led to the recovery of some agro-industrial capacity (CZI 2010). However, often these schemes were alleged to have fiddled with prices, quality and distribution.

Around 2006, 30percent of the A1 land beneficiaries relied on subsidised seed compared to 20percent of A2 beneficiaries, while in high potential districts such as Chipinge 50percent relied on subsidised seeds, compared to nine percent in remote Chiredzi (Moyo et al. 2009). Between 68 and 87 percent of the land beneficiaries, depending on the district, purchased their inputs in markets (Moyo et al. 2009). A marginally larger proportion of A2 compared to A1 farmers benefited from government inputs. Over 96percent of the few land beneficiaries who used agro-chemicals for cotton and tobacco production bought them in markets. Similarly, less than onepercent of the land beneficiaries received government livestock inputs, aside from dipping services. International donors played a limited role in subsidising inputs until 2009. However, while such state subsidies were limited in scale the level of support substantially surpassed that provided during the 1990s.

Formal cattle markets, whose exportation had substantially declined by 2004 due to reduced slaughters and livestock disease, became spatially and socially disaggregated. Meat trade increasingly involves direct sales between producers and new abattoirs, retailers and consumers (Scoones et al. 2010). Livestock production on input sub-contracts and leased grazing created new tied markets (Zvimba field observation 2008).

Government support for grain marketing also increased compared to the ESAP era, but its monopoly became unsustainable. Grain procurement relied on limited subsidies and expensive credit while the grain sold to millers was highly subsidised. Small millers mushroomed in tandem, competing and colluding with established large agribusiness to hike prices (News24 2003). Meanwhile farmers received low grain prices and often delayed payments (FAO 2009). Many producers, consumers, millers and traders circumvented the controlled market leading to high prices in parallel markets and unstable supplies. Few A2 producers ‘obeyed’ the marketing regulations, evoking patriotism while expecting future subsidies (Mhondoro District interview 2009). This limited the potential welfare transfers to deficit areas, despite the substantial fiscal outlays which were supplemented by humanitarian aid.

Around 55percent of land beneficiaries sold their maize to the GMB, while the rest used local markets. Over 22 percent of all the beneficiaries sold their edible beans to the GMB, with slightly more A1 beneficiaries doing so (Moyo et al. 2009). Over 35 and 57percent of the soyabeans produced by the A1 and A2 land beneficiaries respectively were sold to the GMB, which used soyabean trading to generate income, while around 31percent of this crop was retained for own use (Moyo et al. 2009). The rest was sold to agro-processing firms, contractors and local agro-dealers, including some which had secured subsidies. Distance from markets, limited state capacity, and manipulative trading practises by the growing and variegated merchant classes in markets generated various contradictions around state interventions.

Most of the tobacco and cotton was being sold by land beneficiaries to contractors and private buyers in 2006, while over 65 and 55percent sold their sugar and tea in Chiredzi and Chipinge respectively to agro-processing plantations. Between 22 and 40percent of the A2 producers of cotton and tobacco had kept their output that year because prices were poor, while 12 and 24percent of the A1 and A2 farmers could not secure ‘independent’ markets for their sugar, allegedly because buyers ‘condemned’ it (Moyo et al. 2009).

In 2006, the RBZ partially relaxed the policy of withholding large proportions of foreign currency earned by exporters (RBZ 2006). Contract farming benefited from this and became central to the marketing of most commodities, excluding grains, building upon previous experiences with cotton and barley. A few black tobacco merchants piloted such contracting and later sold the businesses to transnationals. The number of tobacco contractors grew to 12 in 2010 from fewer than three in 2003, including four black firms, four new multi-racial contractors and four foreign contractors from China and elsewhere (TIMB 2010). Some of these were subcontractors of Western transnationals such as British American Tobacco. Cotton contractors and buyers also increased and included Indians and Chinese, with the latter up scaling tobacco contracting.

Understanding the role of contract farming in class formation processes is complicated by its unclear association with land sizes and assets owned. Some contract financiers prefer peasants and middle-scale producers because they are less able to resist lower price margins compared to larger-scale producers, who generally have higher social standing and fare better in procuring inputs using their own income, credit and subsidies. Conflicts arose between farmers and contractors over the depressed prices on offer and many ‘resisted’ this by side marketing their contracted commodities to other merchants who had not provided contract inputs to them. But they bought such outputs at marginally higher prices. New sugar outgrowers fought with plantation managers over transportation charges and the pricing of inputs and outputs (The Herald 2011b).

After market liberalisation and dollarisation the limited contract production of foods such as soyabeans faced competition from subsidised imports from the West and genetically modified based grain and oilseeds imports from South Africa. This drove farmgate prices down, reducing local producer incomes. To some extent this propelled many small producers needing cash to reinvest and meet consumption needs into tobacco contracts. This explains the second temporal decline in the soyabean output levels around 2009.

The government revived the Agricultural Marketing Authority (AMA) in 2004 to better regulate markets and introduced new contract farming regulations in 2009 (USAID 2010). By 2011 however the large-scale Zimbabwe Commercial Farmers Union and the CBZ bank, backed by the Ministry of Trade and Commerce and funded by some donors, were establishing Commodity Exchange in Zimbabwe (COMEZ). The AMA feared this would drive speculation on food and increase foreign influence (The Herald 2011c). Class-based struggles over these changing agrarian markets remain potent and open with finance being critical.

Access to finance for farming

Throughout the 2000s the volume of agricultural finance from domestic sources and external concessional loans and aid from Western donors fell sharply compared to provisions during the 1990s (RBZ 2007b). Private agricultural credit declined from over US$315 million in 1998 to about US$6 million in 2008 (MAEMI 2009). European trade credit and the agricultural commodity bonds market had virtually disappeared by 2005. Government credit through Agribank had averaged around US$25 million per annum between 2000 and 2007 and had peaked at US$104 million in 2004, but it declined to below US$3 million in 2007 (MAEMI 2009). Declining revenues limited budgetary allocations and led to larger-scale money printing. This credit constraint fuelled the diversification of the forms and sources of agriculture finance from the mid-2000s.

Subsidised credit, inputs and foreign currency supplies increased (RBZ 2005). Foreign currency was being secured on parallel markets by citizens and government , which fuelled speculative pricing, shortages of goods, and hyperinflation. Around 2004 the government attempted to plug the wheat deficit through subcontracted production and other partnerships with domestic agro-industrial capital on the ARDA estates, using subsidised funds (Moyo 2011b). But this foundered over pricing and profit-sharing disagreements.

Hyperinflation made agricultural credit less ‘competitive’ than short-term trading (Matshe 2004). The World Bank (2006) attributed decreased financing to the undermining of profitability and investment incentives by market controls. Others argued that the land user rights policy and land disputes created uncertainties for investors and limited credit supplies (Richardson 2005, Rukuni et al. 2009).

During 2006, over 78percent of the A1 land beneficiaries used their savings, remittances from home and abroad and non-farm incomes to finance farming operations, compared to 83 percent of the A2 farmers. Only 10percent of them received external financing for production and two percent had access to credit (Moyo et al. 2009). But rising food and input prices after the 2002 drought and global price hikes in 2005 undermined the real incomes of peasants (Wiggins 2005).

Not surprisingly contract farming became central to the financing of smaller and middle-scale farmers (see The Standard 2009) who joined export production to gain access to inputs and increase their earnings. This shifted pre-2000 agrarian relations from the dominance of private credit relationships between large-scale farmers and banks towards bonding more farmers with contracting intermediaries. Before 1986 government had been the major lender (Moyo 1995). When foreign currency and agricultural markets were re-liberalised, agricultural subcontractors escalated such pre-financing arrangements. Private bank credit to agriculture increased to over $300 million in 2010 (MoF 2011), but over 60percent of this went to contractors (USAID 2010).

China played a leading role in financing agriculture through loans for imported fertiliser, agro-chemicals and tractors and in contracting tobacco and cotton from 2006 (Edinger and Burke 2008). By 2009 the state had lured ‘foreign investors’ in partnership with domestic capital to produce and process sugarcane (for ethanol) and increase beef exports, using 20-year build-operate-transfer and land-lease arrangements on parastatal lands (Moyo 2011b). Large agribusiness was regaining dominance in agricultural input and output markets, and new agrarian capital from the East and South was seeking to invest in agricultural exports and to supply inputs. Zimbabwe remained a net exporter of capital to the West during the decade (UNCTAD 2008).

These reconfigurations of the agrarian markets and state interventions fuelled contradictions within the state apparatus over its autonomy and the uneven benefits realised by various farming classes. The re-financialisation of Zimbabwe's economy was increasingly orienting agricultural production to exports but diversifying its global integration. The class dynamics of the emerging agrarian relations reinforced the policy shift that enabled the dominance of large foreign and domestic capital, which is now only peripherally engaged directly in production.

Changing farmer organisation

These shifts in the agrarian relations have altered the political landscape and debates on future agrarian reforms. New forms of farmer organisation and protest are emerging at the local, district and national levels, including among women farmers and through commodity associations (Moyo et al. 2009). Agrarian politics are increasingly shaped by international capital in alliance with new large-scale farmers and the established white and emerging black bourgeoisie around agrarian markets within an increasingly neoliberal regime. Renewed social differentiation among the peasantry and the emergence of a middle class of farmers is being shaped by new forms of accumulation and consumption, which are altering the rural political constituency. Continued sanctions and competition for access to the promised fruits of the indigenisation of mining and other businesses foreshadows new intra-class conflicts, especially among various sections of the elite.

Existing farmers and new settlers mobilise common histories and experiences despite cultural and ethno-regional difference in struggles for land and production resources (Mkodzongi 2011). Church, women's and farmers' clubs and local development associations are being adapted by landholders and others to the new farming areas (Scoones et al. 2010). Local liberation war veterans' and collaborators' associations that led land occupations are recreating their leadership roles (Sadomba 2008). Farm labour activism is being restructured to negotiate labour rights, given their limited protection and the ineffectiveness of the national unions (Chambati and Magaramombe 2008). Former white farmers' unions have been reconstituted. The Commercial Farmers Union (CFU) retains a few members and the new Justice for Agriculture (JAG) mobilises to influence policy and international support over ‘lost’ land and compensation.

Mainstream discourses on the land audit proposed by the Inclusive Government create tensions since multiple landholders and ‘unproductive land users’ could be evicted, however they raise the hopes of some for access to land. Elitist perspectives imply that evictions will affect resource-constrained landholders who cultivate smaller areas (WFLA 2009). Some scholars point to ZANU-PF loyalty and patronage as having dominated land allocations, to the extent that mainly chiefs, civil servants and war veterans benefited (see Zamchiya 2011 in this collection) in parts of some districts, as observed in Chipinge. Yet other studies find that a more broadly based class and social process defined land allocations – and in the nature of beneficiaries – despite the ruling party's overall dominance. (Moyo 2011a, Scoones et al. 2010). This suggests continuing differences over aspects of the land reform but it emphasises the importance of thorough empirical research.

Nonetheless, class struggles over access to land persist since exclusion, tenure insecurities and competing claims over some land and natural resources persist (see Moyo 2011a). The more politically influential and wealthier use administrative fiat, ethno-regional sentiments and at times force to expand their landholdings. Poorer landholders compete mostly through lower intensity confrontations among themselves. Some who did not benefit occupy land and poach resources ‘illegally’. Many local associations seek to defend their land allocations using kinship ties and other sources of solidarity (see Masuko 2008). Local state authorities, including the bureaucracy and traditional leadership, and ZANU-PF structures are central to mediating the social legitimacy of the land redistribution and adjudicating land conflicts.

The state faces pressure to expand agrarian support and market protection from various middle class and capitalist lobbies in farming, agro-industry and trade, including some formed specifically to access subsidies. Over 70percent of the local farmers' associations aggregate their resources to tap public extension and inputs support and to negotiate markets (Moyo et al. 2009, see Murisa 2010), replicating pre-2000 strategies (see Moyo 2002). But middle-scale and larger-scale farmers within the bureaucracy had greater policy influence and received more subsidies (see Moyo and Yeros 2005). Inter-class grievances emerged even among ZANU-PF supporters, as experienced over farm machinery subsidies. Recognising the input shortages facing peasants, the Inclusive Government slightly increased subsidies to them (MoF 2011), while a ZANU-PF presidential ‘well-wishers’ fund also supplied substantial inputs in 2010.

Contract farming relations also fomented the establishment of local farming associations, reinforcing the orchestration of farmer organisation by capital, non-governmental organisations (NGOs), and the bureaucracy seeking access to scattered peasants, as occurred during the 1980s. NGOs that had extensively mobilised small farmer cooperation for rural development retreated during the 2000s (Helliker 2008). Their discourses shifted towards governance and rule of law issues, focusing on the displacement of white farmers and farm workers. Some NGOs however defended land beneficiaries facing eviction (e.g. Zimbabwe Lawyers for Human Rights). Few NGOs advocate the redistribution of remaining large landholdings (Moyo 2011b). But donors and NGOs are consolidating their market-oriented aid schemes for input support for only Communal Areas in collaboration with older farmers unions (see FAO 2011).

New agricultural commodity associations, mostly representing middle-scale farmers, have emerged (Moyo 2011b). For instance, the Sugar Cane Farmers (outgrowers) Association in Chiredzi estates clashed with former white landowner outgrowers and estate managers over services and prices. They protested against the European Union-funded National Sugar Adaptation Strategy, which only finances holders of ‘uncontested’ (i.e. not redistributed) land.

Agrarian politics are also being re-shaped by the changing local administrative and political power relations that resulted from replacing white farmers' control over land, territory and labour. Local influence is now more broadly diffused but the landless are the most vulnerable. Territorial reconfiguration has enabled freer flows of people, goods and services leading to labour mobility, petty trading, gold panning and natural resource extraction. Their regulation is beyond the reach of under-resourced local administrative structures.7 Local power struggles mainly involve lineage-clan leaders, traditional leaders, farmers and social associations, and local bureaucrats, as the powers that large landowners and war veteran leaders of the land occupations had have been replaced. Sparse local government authorities are ill-equipped to regulate the expanded land administration regime and ubiquitous natural resource and mineral extraction. Hereditary chiefs demand more powers to fill these regulation gaps.8

Despite this reconfiguration of agrarian politics some analysts fall back on past discourses shaped by political party polarisation, casting a shadow over progressive politics on egalitarian agrarian reforms. Fortunately most scholars and activists have been coming together to examine the future requirements of agrarian reform, including how to further democratise the land administration system, finance the supply and equitable distribution of inputs and create markets favourable to small producers.

Conclusion

Land reform has restructured Zimbabwe's agrarian relations by reconstituting the agrarian structure, mainly through expanding the numbers of small and middle-scale agricultural producers and reconfiguring the underlying labour relations. The racial and private character of land ownership was largely reversed. While access to land was broadened, large-scale farm holdings and plantations persist with disproportionately more land than is warranted, and this sustains continued landlessness. Agro-industrial estates were marginally restructured by introducing more small-scale outgrowers and through the expansion of public estate production, increasingly in partnership with foreign capital. Pressures to commodify land persist as various elite classes demand the reinstitution of freehold tenure and pursue informal land rentals. While agrarian labour relations are now more dominated by self-employment in diverse farming and non-farm activities, and part-time wage-labour is more common, prevailing wages and incomes remain repressed by low productivity, exploitative commodity markets and slow recovery of production in other economic sectors.

Heterodox interventionist policies introduced from 2002 initially shaped changes in the agrarian markets, including uneven access to subsidised inputs, shortages of goods and hyperinflation. Overall agricultural output had declined, but this began to rise slowly and selectively from 2006. A broader range of producers invested in cultivating much more land mainly for food, although land productivity remained lower than is potentially achievable due to limited access to farming inputs in markets, despite the subsidies. Many more producers are now involved in the production of diverse exports than was the case before 2000, suggesting a degree of income redistribution. However export commodity producers have the greatest access to farming inputs and commodity markets, and such access generally favours larger landholders who can also hire more labour and accumulate more assets.

The supply of farming inputs and credit and the buying of commodities by established agro-industrial conglomerates, merchants, and banks, which had been substantially reduced by 2007, was generally propped up by state subsidies. Input supplies were revamped by the re-liberalisation of markets in 2009. Agricultural production is increasingly being organised through contract farming, focusing on export commodities. Chinese contractors and loans negotiated with the state fuelled the revival and reconfiguration of agrarian markets and lured back capital from the West.

This restructuring of Zimbabwe's agrarian relations has the potential to deepen the autonomy of the peasantry and intensify productivity towards increasing supplies of more nutritious foods and raw materials for the home market. But these gains are limited by the increased export orientation and ascendance of agribusiness over popular markets, limited fiscal capacity and the statist defences against sanctions. While class struggles over land, agrarian markets and labour now entail new and more broadly based forms of smaller producer organisations and protests, these are pitted against an emerging alliance of larger-scale farm producers and foreign monopoly capital linked to expanding multi-racial domestic capital, whose membership traverses the political party lines.

Progressive agrarian change should consolidate the trajectory of smallholder development by securing peasants' control of land and protecting their markets to enhance popular food sovereignty, productivity and cash-earnings from self-employment in farming and non-farm industry. This requires stronger producer and consumer cooperation against agri-business and contractors (from the West, East and South) that link input and credit supply to crop purchases at depressed prices and against trade policies which enhance the dumping of subsidised foods. Expanding the democratic space and regaining policy autonomy towards articulated national development will be critical to the agrarian politics of advancing the peasantry and industrial diversification. Redirecting the expanding rents from mining to support this agenda will be crucial. As elsewhere, agrarian change under contemporary imperialism is neither linear nor even.

Additional information

Notes on contributors

Sam Moyo

I thank Ndabezinhle Nyoni, Paris Yeros, Lionel Cliffe and four anonymous reviewers for their comments.

Notes

¹For details on the districts, which have diverse agro-ecological potentials, mixed farming systems and ethno-regional contexts, see Moyo et al. (2009, 8–13).

²These include Chipinge, Chiredzi, Goromonzi, Kwekwe, Mangwe and Zvimba. The survey was accompanied by key informant interviews, group discussions, field observations and other open interviews. Secondary evidence and information gathered during two land audits (Buka 2002, Utete 2003) complement the survey data.

³The A1 model targeted landless and poor families, providing land use permits on small plots for residence cropping and common grazing, while the A2 scheme targeted new ‘commercial’ farmers, providing larger individual plots on long-term leases to beneficiaries supposedly with farming skills and/or resources (including for hiring managers).

⁴ Statutory Instrument 235A of 16 July 2001. Grain Marketing (Controlled Products Declaration) (Maize and Wheat) Notice, 2001 in terms of section 29 of the Grain Marketing Act (Chapter 18:14).

⁵Personal communication, 2010.

⁶Interview with Theo Khumalo of Colcom, April 2011, Harare.

⁷The state has tried but failed to stop popular exploitation of such resources using environmental and mining regulations.

⁸See statement by Chief Charumbira at a COPAC meeting held in Harare, on 4 January 2010.