How and why large scale agricultural land investments do not create long-term employment benefits_ A critique of the ‘state’ of labour regulations in Ghana

Support for large scale agricultural investments in Africa has been mainly premised on their employment prospects for local populations. However, despite earlier calls by Tania Li to centre labour in the land grabs debate, labour is generally invisible in both mainstream policy and academic research. This paper, through a governance lens, draws attention to the implications of the global land rush on wage labour. In principle, policy frameworks that emphasise the labour potentials from large-scale land investments also gravitate towards regulations that seek to facilitate capital accumulation and mitigate negative impacts on communities – congruent with Ghana’s policy direction. This paper assesses the political-economic context of the legislative gaps in the current governance framework for wage labour and large-scale agriculture in Ghana; characterised mainly by absent, illusively present and repressive institutions. It is supported with empirical findings from the nature of farm workers’ incorporation into a transnational oil palm plantation in Ghana, their struggles over the nature of the investment, and the political orientation of the existing regulatory institutions. The study calls for policy measures which address power relations that shape the distribution of benefits from land investments, and also recognise structural inequalities that exist in and outside of agriculture.


Introduction
The past decade has seen an overwhelming amount of literature on the impacts of transnational agricultural investments in Africa. Much of the debates have been rooted in theoretical and analytical perspectives that speak particularly to the land question, leaving labour to its margins (Gyapong, 2020;Moyo et al., 2012). Thus the many excellent discussions on land markets, tenure security and property rights to govern capitalist agriculture (see Aha and Ayitey, 2017;Ayelazuno, 2019;Byerlee and Deininger, 2013). Many studies have also criticised large scale investments for the negative livelihood implications of dispossession (Ampadu, 2013;La Via Campesina, 2013;Levien, 2011;Margulis et al., 2013aMargulis et al., , 2013bPatnaik and Moyo, 2011;White et al., 2012). When land deals occur, modifications to land access are visible in the short term, while labour related impacts may only materialise at later stages (Cotula et al., 2014) Additionally, full implementation of land investments occurs at later periods making it methodologically challenging to study (Oya, 2013a, Oya, 2013. Even so, with many land deals in their implementation stages, it is essential to assess labour impacts. It is not to say that land and labour issues can be, or are researched as entirely distinct aspects of land deals; instead, their apparent nexus has not received the necessary attention. Li (2011) admonished that we investigate what translates on the ground when investors need both land and labour. A few studies have begun to track the employment implications of large-scale agriculture. Different agricultural models have different labour implications (Hall et al., 2017); there are accounts of skill and gendered income distribution and negative health impacts on farmworkers (O'Laughlin, 2017). Some researchers have reported that regardless of the often low wages, farmworkers are still attracted to the security of income flows that guarantees their access to basic needs (Schoneveld et al., 2011). Most of the emerging evidence suggest a complexity of issues regarding the amount of jobs generated, the differentiated inclusion opportnities and job insecurity (Dubb et al., 2016;Gyapong, 2019;Hall et al., 2017;Kay, 2015;Ouma, 2018).
How then should large-scale agricultural investments to be governed? Borras et al., (2013a) mapped out three broad competing policy directions regarding the forms of regulations that are deemed appropriate and the purpose for which they are to be maintained. They are policies that regulate to facilitate; regulate to mitigate negative impacts; regulate to stop and roll back large-scale agricultural investments. In Ghana, agriculture plays a key role in the development, and successive governments have sought to modernise the traditionally predominant small holding systems by integrating them into the markets of largescale schemes and the creation of an enabling an environment for private investments (Amanor, 2010). There has been longstanding state support for plantations in Ghana. Examples include the post-independence compulsory land acquisitions for oil palm and rubber plantations-many of which are now privatised-to the promotion of foreign direct investments in agriculture through market-based large scale customary land acquisitions in contemporary times (Amanor, 1999;Tsikata and Yaro, 2011). For instance, in 2012, the Ghana Commercial Agricultural Project (GCAP) was launched with the objective of 'improving the investment climate for agri-business and developing inclusive Private-Public Partnerships (PPPs) and smallholder linkages 1 '. To emphasise this, 'modernizing agriculture' remains the overarching theme of Ghana's agricultural policy as well as the new private sector development strategy. It focuses a substantial role for the private sector in transforming agriculture from a low-productivity subsistence-based sector to one characterised by high-productivity, integrated value chains, and extensive value addition. Government interventions are focused on the enabling environment and other targeted measures to facilitate such investment.
Similarly, in 2015 the ministry of food and agriculture made available a list of public lands and priority crops earmarked for largescale investments while inviting partnership opportunities from industrialised countries. Many times, these ideas also resonate with traditional authorities 2 who premise their expectations on jobs for the youth. These have also been accompanied with measures aimed at ensuring transparency and inclusiveness of local populations. Examples include the state's strengthening of land administration through registration, and certification, and the voluntary land investment guideline formulated by the Ministry of Food and Agriculture in 2015to pre-empt and address challenges with land investments.
Generally, Ghana appears to incline towards the first two regulatory tendencies of promoting land deals, but what does this mean for rural working people affected by land deals? This study aimed to investigate the effect of regulations intended to ensure responsible investments on labour outcomes for local people, in particular, marginalised groups. The study explored the experiences of farmworkers on an oil palm plantation in Ghana and the capacity of existing land and labour regulations to ensure equitable benefits from land investments. It is important to note that the goal of this paper is not to present detailed empirical evidence of wage labour relations on the oil palm plantation. Instead, it is a critique of the existing regulatory environment for governing wage labour and investments, while reflecting on the experiences of the farmworkers in order to ground the discussion in firsthand observations.

The politics of regulating to promote large-scale agricultural investments
The tremendous media and research attention on large-scale agricultural land acquisitions have contributed to a diversity of governance strategies by policymakers, civil society organisations and researchers; and at different political spaces-global, national and local levels. Policies are also being shaped by debates that are shifting from simplistic accounts of good (people) versus bad (investors) to more critical and complex analyses of issues (Byerlee and Deininger, 2013;FAO, 2012;Fao, 2010;Margulis and Porter, 2013). Indeed one would hardly find any fixed governance approaches, however, the three competing tendencies: (a) regulate to facilitate land deals; (b) regulate to mitigate negative impacts and maximize opportunities; and (c) regulate to stop and rollback land grabbing, put forward by Borras et al. (2013) is an appropriate starting point to understanding to the institutional aspects of large-scale agricultural investments in Africa. Reiterating the caution by Borras et al. (2013 pg168), 'these tendencies are not sharply defined and fixed. Hence the use of the term 'tendency', and each is in turn, internally variegated' and also influenced by broader political and ideological contestations around the notions of economic growth and development. These three regulatory approaches help to analyse the policy directions of states while recognising the political shifts overtime.
The tendency (A) of regulating to facilitate land deals is embedded in administrative propositions to address the issues of effectiveness, efficiency and transparency in land transactions (Borras et al., 2013;Franco and Monsalve Suárez, 2018). Here, there is the recognition that large scale agricultural investments are driven by the global demand for food, feed and fuel. However, they are also considered favourable because of longstanding advantages, e.g. market adaptability and economies of scale of large operations (Deininger and Byerlee, 2012). This policy direction facilitates accumulation under transparent and participatory institutions which checks on information asymmetries. Proponents promote land formalization programmes, public-private partnerships and property rights purported to secure investments and reduce transaction costs for investors. They are also presumed preconditions necessary to prevent failed deals, unemployment, and irreversible environmental implications (UNCTAD et al., 2010).
The tendency (B) of 'regulating to mitigate negative impacts and maximise opportunities' recognises the inevitability of large land investments due to communities' desire for investments amidst fiscal limitations and the lack of trust in the state to spearhead such developments. Similar to tendency A, policy priorities emphasise the need to ensure that investments are undertaken sustainably-have minimal adverse effects on people and the environment (von Braun and Meinzen-Dick, 2009). However, unlike the 'strategic thinking underpinning the first tendency which advances investments, this second current adopts a more tactical approach-to address the urgent needs of poor people whose livelihoods could be threatened by land deals (Borras et al., 2008;La Via Campesina, 2007). At a period when the land rush literature was still primarily based on speculations, (Behrman et al., 2012 pg. 71) suggested that 'If large-scale land investments are properly executed with appropriate attention to gender dimensions, land deals can provide transformative opopportunities for both women and men through the introduction of new employment and income generation opportunities, new technologies, and new services. Appropriately designed land deals that include local women, as well as men in the negotiations and compensation, may even aid in the distribution of local resources in a more gender-equitable fashion. Investors also stand to benefit from such gender-equitable land deals that consider the full range of skills, labour potential, and knowledge of local women and men' In the past few years, human rights advocacy including the call for decent labour standards (e.g. elimination of child labour, right to unionisation, social security benefits, generation of new and higherpaying jobs), land use rights, transparency, and the incorporation of small-scale farmers in production and market chains of agribusinessese.g. out-grower schemes, have all become prominent ways to address land grabs (Beekmans et al., 2014;Bertram, 2012).
In stark contrast to the first two approaches, is the tendency (C) of 'regulating to stop and roll back' land deals. It represents a radical stance against large scale investments. From this perspective, land deals expose rural people and peasants to dispossession and forced proletarianization, eventually culminating in deprivation and poverty (La Via Campesina, 2007;Moyo et al., 2012;Rosset, 2009). McMichael (2008) adds that beyond dispossession, is an equally crucial fact that such investments tend to rule out the place for peasants, closing doors to alternatives and re-moving peasants and their systems of production from history. Similarly, others highlight their embeddedness in global power asymmetries whereby investors as compared to local populations, become better positioned to advance their agenda and benefit more under institutions of property rights (Margulis et al., 2013a(Margulis et al., , 2013b. Notably, radical agrarian justice organisations such as the food sovereignty movement has been influential in proposing alternatives through campaigns for agrarian reforms that support peasants agriculture instead of large scale investments-a popular declaration of La via Campesina has been "Our Land is our identity, it is not for Sale.' (Borras and Franco, 2012;La Via Campesina, 2007) While all three tendencies recognise power asymmetries as a constraint to the developmental outcomes of land deals, they diverge on the perceived sources of these power inequalities, and how to deal with them. Whereas policy strategies that navigate towards the first two tendencies show optimism that inclusive participation and transparency could lead to positive outcomes, the latter stresses more on the expropriating and coercive character of capitalist development which makes it unlikely to trust participatory processes. Again, from outside and within the third regulatory tendency, class, identity and ideological tensions persist within radical agrarian and environmental movements and the extent to which they represent the diverse interests of rural people or peasants in different contexts also remains highly debated (Bernstein, 2013;Edelman and Borras, 2016). A project may be denounced globally and nationally but receive support from some groups at the local level (Larder, 2015). For instance, contrary to the radical demands of agrarian movements, in the Ghana case studied in this paper, there is a persistent investment and employment desire that supersedes a land grab narrative.
These regulatory responses, therefore, do not necessarily lead to particular outcomes but requires evaluating what happens on the ground concerning the specific political-economic context within which they are implemented. Thus, it crucial to examine the political processes of certain regulations and why they emphasise certain goals than others (Steinmo et al., 1992). O'Connor (1973) argued that states have a double imperative of maintaining power by promoting economic growth through capital accumulation while at the same time, safeguarding their legitimacy. It is thus not surprising that in many African states, including Ghana, the first two political tendencies are prevalent. As illustrated earlier, successive governments in Ghana have been promoting agro-industrialisation, foreign direct investments and agricultural technologies to 'transform' smallholder production and commercial-scale schemes for local development, thereby making largescale land investments attractive to the state (Teye et al., 2016). Several land deals in Ghana have been supported and justified by chiefs on account of securing jobs for the youth and providing necessary socioeconomic infrastructure in deprived rural communities (Ahmed et al., 2018;Boamah, 2014). Recent land transactions have been primarily between traditional leaders and investors, with the state playing a secondary or covert role. Nevertheless, tensions often exist in the autonomy and capacity of the state and traditional authorities to balance these responsibilities to benefit different groups equitably (Fox, 2000).
In many instances, institutions operate within high power imbalances (see Visser and Spoor, 2011). For instance, in Ghana, the extent to which people can hold the state accountable on land deals depends on the land tenure system guiding the acquisition. Ghana's policy of 'non-interference' in customary and market-based land tenure (Amanor, 2008) promotes capital accumulation without profound legitimation pressures-revenue for the state, while investors also operate under a laissez-faire business environment. Even though the state's nondirect interference in customary land transactions empowers traditional land institutions, it also relieves the state from excessive legitimation pressures thereby complicating the balance of power in land transactions and the ability to mitigate negative livelihood implications, e.g. labour impacts.
The labour question draws attention to issues of livelihoods but also the classical problem of proletarianzation, i.e. forced labour driven by land dispossession (Marx, 1977;Scoones, 2009;). To adequately address the labour question requires a balancing of analytical perspectives around issues of agency and structure. In other words, the orientation of regulations/institutions, the business models of investors, as well as the desires of affected communitiesare all socially and politically constructed, and therefore should be seen as broad patterns of class struggles around capitalist development in the countryside (O'Laughlin, 2002;Steinmo and Thelen, 2010). Labour casualisation, for instance, is often associated with insecurity and low income (Yaro et al., 2017). However, for those who consider it as a potential livelihood opportunity, they emphasise the benefits of flexibility in allowing workers to engage in more than one productive sector and even the possibilities of technology transfer to small scale farms (Deininger and Xia, 2016). At the same time, it could enable investors to hire and fire workers with ease, make more elastic use of skills, introduce non-conventional work arrangements and lower wages according to their business needs and performance in ways that may not necessarily benefit workers (Arnold and Bongiovi, 2013, p. 6). In effect, the struggles over incorporation imply struggles over the character and orientation of labour institutions and models of production, all of which affect the outcomes of and responses to land deals in both expected and unpredictable ways. The study is, therefore situated in the interactions between class and livelihood struggles of farmworkers in a regulatory environment characterised by structural inequalities.

Study approach and the main sources of data
One of the two main sources of data for this study is the labour related experiences among workers on the Volta Red Oil palm plantation in Ghana. It is affiliated with the infamous American owned Sithe-Global Sustainable Oils (SGSO), popularly also known as Herakles Farms, also operating in Cameroun. It started in 2008 as a pilot project involving a 50 year lease of approximately 3750 ha of family lands (semi-deciduous forests) belonging to 15 extended families (constituting hundreds of family members) of the Ntrubo clan in the Nkwanta South Municipality. In 2009, the company acquired a twenty times bigger land size in Cameroun (The Oakland Institute, 2016). Not long after, the land-owning families in Ghana began litigation on the lease agreement, and thus it did not make economic sense for the company to be at the forefront of the project 3 -they concentrated their attention to the Cameroon investment. Since 2013, the day to day operations of the plantation has been managed by Ghanaians but under the directorship of British investors 4 of the Wyse Group UK 5 who had already established their presence in the oil palm sector in the adjoining Kadjebi District as Volta Red Company limitedan oil palm processing mill that was at the time, dependent on their 41 ha oil palm plantation and an out-grower scheme 6 .
Field reports indicate that since 2013, Volta Red took over the Herakles farms/SGSO in Ghana and managing it concurrently with its original investment. 7 The target market for the product -oil palm -is the local and the neighbouring West African markets as according to the management. As a company, Volta Red employs approximately 300-350 people on the three investments-two plantations and the oil palm processing mill. The study focuses on the approximately 250 8 . 3 Interviews with the Project Liaison and landowners in 2018 4 resident 5 https://suite.endole.co.uk/insight/company/05665971-wyse-group-limited 6 From field interviews 7 Some reports indicate that director has taken over management of the contested Herakles project in Cameroun see (https://www.greenpeace.org/ africa/en/blogs/1525/the-new-name-behind-the-threat-to-cameroons-forests/) 8 This is an upper limit from a workers list of 237, during my fieldwork I did not get the actual numbers because of the casual and seasonal nature of the work. In essence, some casual workers have contracts, and others do not workers at the Fankyenekor-Brewaniase plantation of 3715 ha in Fankyenekor Brewaniase, previously owned by Herakles Farms. A quarter of the workforce are migrant labourers from across the country and neighbouring Togo who moved specifically to work on the plantation, while the rest commute from the three main catchment areas around the planation-Brewaniase, Abuburuwa and Dodo-Tamale. I conducted field visits in Ghana between February 2018 and March 2019. I adopted a qualitative-dominant mixed methods approach, conducting a socio-economic and employment survey with 200 farmworkers, informal conversations and interviews with random community members, former employees, and farm workers (80 people). I also carried out gender and task-based focus group discussions with the workers and made use on my physical access to the farm to observe their working conditions. Other key informant interviews were conducted with the management, seven out of the eight supervisors, the land-owning families and their family heads, traditional authority, and (local) government officials, officers of the Ghana Agricultural Workers Union (GAWU). Between October and December 2018, I gathered and reviewed national (Ghana) policy documents including the constitution, the Labour Act, 651 of 2003, as well as other (inter)national soft laws and the voluntary guidelines on land deals. The findings from the policy and legal documents are central to the discussions in the subsequent sections.
As an overview, several institutions govern labour in Ghana. The Constitution serves as the main legal umbrella for employment-safeguarding the right to work; protecting the health, safety and welfare of all persons in employment; and setting some fundamental working conditions such as equal pay, limited working hours and holidays. The constitution guarantees workers' rights to unionization and prohibits forced labour. Specifically, the Labour Act, 2003 (Act 651), and the Labour Regulations, 2007 (LI 1833) stipulate the legal framework for employment in line with the constitution of Ghana. It addresses the scope for the protection, conditions of employment, and unionisation while dedicating sections to women, youth and persons with disabilities. Part (X) of the labour Act, 2003 focuses on casual and temporary workers (Republic of Ghana, 2003). Here, the act emphasizes the 'fair' and non-discriminatory practices of remuneration. There are other complementary laws such as the Persons with Disability Act 2006 (Act 715), the Workmen's Compensation Act, 1987 (PNDCL 187), The National Pensions Act, 2008 (Act 766) and the Internal Revenue Act, 2000 (Act 592), all of which should in principle, enhance the administration the labour markets.
The Ministry of Employment and Labour Relations serves as the overarching body for governing labour. In 2015, it formulated the national employment policy (Republic of Ghana, 2015b) in line with the International Labour Organisation (ILO) standards. The policy addresses four core issues of increasing employment generation, quality of employment, labour productivity, and enhancing governance of labour markets. It has decentralised and devolved departments and commissions responsible for the day to day administration of labour at the local level. In 2015, the Ministry of Food and Agriculture also designed the Community/Investor Guidelines for Large-Scale Land Transactions. The guideline sets a framework for cooperation among investors, communities, and government and suggest their roles in the five stages of investment: preparation, initial engagement of stakeholders, impact assessment, negotiation and consent, and monitoring and enforcement of agreements (Republic of Ghana, 2015a). In principle, governance of the agricultural workforce in Ghana draws on the above regulatory and institutional frameworks, yet to what extent are they able to adequately secure the rights of farmworkers?
2. Job creation rhetoric under a laissez-faire investment environment I have indicated earlier that one of the key assumptions for the promotion of farmland investments in Africa and other developing regions is the employment generation potentials. By this, policymakers put faith in institutions to ensure that agribusinesses provide jobs in the localities within which they operate. There are, however, several challenges with this assumption, especially within the Ghanaian context. First, there are no hard laws or national regulations that specifically enforce employment creation on transnational large scale farms. During my fieldwork, the failed promises of employment was the biggest disappointment for the affected families and communities. This situation reflects the non-concurrence between verbal promises of investors and the actual content of the lease agreement 9 bearing the the stamp of the land registry. Per the lease document, the survey, and interviews with families, the company is upholding its stated obligations of employing at least one affected family member provided they have the requisite skills and qualifications. At least thirty-five (35) of the workers identified as direct proletariats, but this represents a very low percentage of the hundreds of resident nuclear families that constitute the fifteen extended families affected by the land deal. Likewise, most of the clauses in the lease document are vague and non-binding on the company. An example is one that ties their social responsibility to a condition of being able to acquire and plant about 10,000 more hectares of oil palm in the Volta region of Ghana. However, even at the time of the study (10 years after concession), the company was only maintaining just about a 1000 ha (less than a third) of its existing concessionso in principle, the community cannot hold them accountable. It appeared to be struggling to stay in business, and some workers confirmed that in the previous years, it was at the verge of collapsesome pointed political-ecological reasons similar to the incidences of failed farm projects in Ghana (Ahmed et al., 2017;Tsikata and Yaro, 2014;Yaro and Tsikata, 2013). Thus, not only have expectations of wage employment been unmet, but also the hopes of indirect jobs.
Besides, the existing regulatory institutions in Ghana do not have the autonomy and capacity to dictate the business model of investors while at the same time, the Free Zone Act (1995) provides extensive and generous financial incentives to foreign investors. 10 Although certain regulations like that regarding environmental impact assessments and land registration/certification could be guaranteed, it falls outside the priorities or jurisdiction of state institutions to determine and monitor how they run their businesses-be it labour intensive, mechanised, out-grower schemes or even the types of crop to be cultivated. Hall et al. (2017), provide rich insights into the emerging patterns and trends of the labour implications of different models, highlighting the diverse opportunities and risks for different social groups. For instance, the extent of mechanisation or labour intensity may depend on several factors, but primarily, investors would not compromise on the financial returns of their businesses. Promoters of oil palm plantations often emphasise job creation, yet oil palm requires an average of one worker per five hectares which is lower than other crop types such rubber (World Bank and International Finance Corporation, 2011).
Although large-scale farms could have potentials for extensive job creation, economic viability and 'social sustainability' do not necessarily cohere, but no one seems to be responsible for this gap in logic. Critics of land deals often emphasize the perils of surplus labour resulting from redundancy. However, this case shows an even more complicated situation: the plantation work is labour intensive and needs workers to be able to maintain the remaining 2500 ha, yet management is compelled to cut down cost to remain in business; people (especially men) want jobs, but the poor conditions deter them; women have minimal job opportunities; there are high rates of absenteeism from casual workers, yet without commensurate sanctions due to labour supply constraints; management seems to be planning 9 . They were written by a private legal firm in Ghana for the company without the involvement of the land-owning families 10 See https://gfzb.gov.gh/index.php/incentives/ towards increased mechanisation in the long term, which could further reduce labour opportunities. The voluntary guidelines are well intended to address or pre-empt irresponsible investments, but unfortunately, in many cases, such soft laws are already too late and not binding. Many communities have already become victims to contentious land acquisition processes that may require extreme and specialised interventions to revert their process or their implications. In the Herakles-Volta Red land deal, the initial processes of consultation, and the content of the lease agreemente.g. a four -year moratorium on the payment of rents, and absence of binding clauses on employment creation raise many questions. Yet, the general expression from the affected-landowners is that akin to a spilt milk: they are bound to the existing lease agreement prepared by lawyers of the company, and family heads are waiting for its 50 year expiration date to take necessary actionsa situation which the youth hope to change sooner than later, through organised action. Moreover, the guidelines are vague, issues on labour are less discussed as compared to land tenure, and their recommendations pay more attention to participatory consultations rather than power dynamics. For instance, the Community/Investor Guidelines for Large-Scale Land Transactions in Ghana (hereafter the Ghana Guideline) treats communities, investors and the government as though they are stakeholders with equal interests and power. In the guideline, A proposed best practice is for investors to create a sequential plan for business development, clearly identifying the timing of the sequential phases, the amounts and types of land needed in each phase, as well as the expected crops, returns, and employment projected in each phase.
Doing so should help create win-win possibilities for both investors and communities. That is, communities can then ascertain the long term demand for land in order to avoid initial acquisitions that are larger than necessary (Republic of Ghana, 2015a, p. 8).
In an interview with the affected families, almost all of them expressed regret to agreeing to the land transfer, but their major worry has not been because of dispossession 11 rather, the failed promises of decent jobs. Although the Ghana guideline recommends employment creation as a component of every large-scale agricultural investment, it is discussed from the angle of compensation and non-monetary benefits that should accompany land deals, rather than being pursued as an 'obligation' of the investors' or a 'right' to the affected communities. Such propositions do not necessarily guarantee that recommendations are upheld by investors nor may they even be feasible in the context of many developing countries characterised by political, social and economic instabilities. Indeed, there are differences among investors depending on the scale, extent and duration of establishment, the models of production and tenure agreements. However, these individualised processes leave the livelihood needs of marginalised communities to the 'goodwill' of investors.

Agricultural wage labour and the institutional bypasses
Transcending the soft guidelines to the well-institutionalized laws and regulations that guide employment relations, where is the place of wage farmworkers? A critical gap in the existing labour institutions in Ghana is that they do not have binding responsibilities in the agricultural sector. For instance, the Department of Factories Inspectorates Division obtains its mandate from the Factories, Offices and Shops Act, 1979 (Act 328), obviously excluding agricultural work. The National Labour Commission, which was birthed by the enactment of the Labour Act, 651 of 2003, has as its mission., 'to develop and sustain a peaceful and harmonious industrial relations environment through the use of effective dispute resolution practices within the context of the law, promotion of cooperation among the labour market players and mutual respect..' Hence, in the events employer-employee disputes on plantations, parties cannot rely on the labour commission to perform its core function of addressing complaints. . Although per the mandate of the labour department, general enforcement of labour standards apply to the agriculture sector, their scope of operation is too broad and allencompassing (Akorsu, 2010). Thus, it came as no surprise after interviews at the local government institutions, that they knew very little about the operations on the plantation, except the company's tax commitments. The national labour laws and regulations have also been designed to be primarily applicable to industries and services. The 2003 labour Act (Act 651) emphatically eliminates agricultural workers from the coverage of the law even though they happen to be among the most marginalized groups in the country. Under the section on 'special provisions relating to temporary workers and casual workers', the Act does not only prioritize investors over workers -'employer may hire a worker on terms that suit the operations of the enterprise', but also states that the 'section does not apply to, pieceworkers, part-time workers, sharecroppers, sea-going personnel in the fishing industry who are wage earners'.
Labour on the Herakles-Volta Red plantation is characterized by piece rate, labour mobility across tasks, seasonal layoffs, six-month casual contracts with no guaranteed progression, and casual workers. About seventy per cent (70%) of the labour force are casual and temporary workers and approximately nine out of every ten women are casual workers. Conscious of their insecurity and the lack of appropriate regulatory interventions, the workers diversify their livelihoods with other occupations. More than eighty per cent (80%) of the workers were engaged in what they considered as equally important activities such as petty commodity trading, farming and transport services. Many of them do these additional jobs not because casual work is inherently favourable to them-it is far below their knowledge of decent employment.
Again, in Ghana, occupational health and safety of workers are under the Workmen's Compensation Law 1987 (PNDC 187), and the Factories, Offices and Shops Act, 1979 (Act 328) but these also fall short in addressing farmworkers issues. The former is a discretionary directive which is also more favourable to employers than workers. As the name suggests, the latter prioritises industry, thus limited in ensuring the health and safety for workers in the agricultural sector-including farm, fisheries and forestry (Rep. of Ghana, 2014). Meanwhile, plantation farmworkers are often exposed to diverse forms of injuries, accidents and health risks that go unaccounted for in these laws. Most large-scale production schemes in Ghana rely heavily on chemicals that are manually applied by farmworkers. For example, chemical sprayers expressed concern about inadequate protective clothing and irregular health checks. Women who engage in the picking of loose palm fruits with bare hands are also susceptible to nail infections and reptile attacks, while many harvesters and pruners also complained of chronic chest and neck pains. Casual (semi) skilled workers complained about the risks of operating faulty machinery and without licences. The reality is that, although many of these farmworkers double as own-farmers, the plantation-style of production requires continuous learning which sometimes takes a while to master. In a study on sugar cane plantations in Mozambique, O'Laughlin (2017), found similar incidences whereby the intensification of work exposed farmworkers to severe dehydration, eye infections and respiratory difficulties and long-term health conditions like chronic kidney disease associated with dehydration. In effect, the agricultural sector 12 , which employs the largest share (44.7%) 13 of 11 with the exception two extended families that lost all their land to the acquisition 12 Including forestry and fisheries the total labour force in Ghana is not fully represented in national legislations.
In the few policy circles where labour regulations and institutions apply to farmworkers, institutions struggle with fragmented and overlapping functions and mandates. Since 1985, when the Department of Factories Inspectorate was removed from the Labour Department, there has not been any clear distinction in their responsibilities. Regarding issues of occupational health and safety, other government institutions such as the Ministry of Health, Environmental Protection Agency, and Fire Service have several overlapping mandates in terms monitoring (Akorsu, 2010). Under such circumstances, there could be either duplication of efforts, or no job done. Other practical and technical challenges hinder the ability of labour institutions to function effectively. The various labour organisations are inadequately staffed and do not have the necessary logistical support to carry out inspections and enforcement of standards and labour market research (Akorsu, 2010;Oya, 2013a, Oya, 2013. In a study of the Indonesia Palm Oil sector, Li (2017) also demonstrate how sometimes companies can evade unfavourable reports through their alliances with politicians and other bureaucrats; a situation that resonates in the Ghanaian context. It is for such reasons that de Schutter (2011) cautions against an overestimation of the capacity of the governance structures to facilitate land deals for local development.

Institutional legitimation of status quo and repression
De Schutter (2011, p. 258) argued that research that investigates whether land investments represent an improvement from the status quo ante do not provide a holistic understanding of impacts. Policymakers and researchers need to go beyond here-and-now-narratives to ask if rural lands could not be used more productively, equitably and sustainably under structural, institutional and agrarian reforms that distribute land to smallholders. Akram-Lodhi (2007) and Whitehead and Tsikata (2003), also have questioned 'win-win' assumptions in the promulgation of formal land administration programmes such as titling, certifications and land registries and how these regulations usually perpetuate existing inequalities in land access and control. In a similar regard, the issue of labour also calls for attention, yet mostly underrepresented in these debates. The agrarian question of labourwho does what, and who gets what, in relation to the social divisions of labour and the distribution of income are fundamental issues for understanding capitalist development in the countryside (Bernstein, 2010). Asking these questions help unpack how people are integrated and how dynamics produce social forces with different political possibilities (Bernstein and Pitt, 1974, p. 522;Henry. Bernstein, 2010).
To some extent, the income of the farmworkers is partially 14 guided by the minimum wage regulations in Ghana. The introduction of minimum wage dates back to the 1950s when Ghana became independent from British Colonial rule. Since the enactment of the labour Act 2003 (Act 651), the payment of minimum wage is binding on employers even though in practice not many uphold it, particularly in the private and informal sectors (Kumah, 2017). During the economic crises in the period between 1970 and 1990, where cost-cutting structural adjustment programmes advanced, so was the rise in debates on the distortion effects of minimum wage on an otherwise efficient labour markets. Once again, the minimum wage has grabbed policy attention, but now as an effective measure to address poverty and as a means of social protection for vulnerable groups. The assumption is that it provides a guarantee for low-wage-low-skill workers to earn "decent" wages and sustain at least a subsistence standard of living. All the same, others are sceptical about its effects on real income and employment in many developing country contexts (Obeng, 2015, p. 86).
In the past three years, the minimum wage in Ghana increased annually by 10 per cent, yet this represents a rise from just GH₵8 to GH ₵10.6 per day (approx. 1.5 to 2 USD 15 ), which is woefully low to even maintain a subsistence living as compared to many developed countries where households can afford decent living with minimum wage earnings. Evidence from the study on the Herakles-Volta Red oil palm plantation showed that even though the farmworkers were being paid almost fifty per cent higher (GH₵14.04) than the 2018 national minimum wage of GH₵ 9.68, it was still challenging for farmworkers to maintain decent living for themselves and their households. Of course, per the organisation of labour (piece rate) and tasks on the plantation, the impacts are differentiated among the working groups -men, women, the ageing, landed and the less-landed. Nonetheless, even for those who are better positioned to work long hours or engage in lucrative labouring tasks on the plantation, high incomes are accompanied by negative trade-offs-especially on their healthcare. These centralised wage regulations do not reflect the social and cultural dynamics of rural spaces with large household sizes, extended family responsibilities and educational aspirations that force farmers into becoming semi-proletariats.
It is worth noting that most of the peasant farmworkers 16 did not deny the benefits of having access to regular income when compared to their earnings from their peasant farming. Many of the male workers interviewed indicated their dependence on wage income to invest (in labour and agrochemical inputs) and possibly to expand in their own farms, as well as to provide shelter (mostly traditional mud houses) for their households. Wage income also means a lot to women, most of whom have become the 'duty bearers' of their children's education. At the same time, many male youths considered themselves temporary workers, saving their income to pursue higher education with the hope of future work in the industry or service sector. However, not only do these well-being benefits from the meagre wages demonstrate the poor initial conditions of the workers, but it also shows how benefits are usually unequally distributed among different groups. A review of the attendance rates on the plantation showed that women farmworkers appear to be more dedicated to their jobs than men, yet ninety per cent (90%) of them are casual workers, and they receive the lowest wages (See Baumgartner et al., 2015 on their economic analyses of employment outcomes in Ethiopia). It was therefore not surprising to find a high prevalence of debts among the workers (40 % of them). The expanded cash economy has also increased borrowing -casual workers hope to migrate to permanent contracts mainly to have access to bank loans 17 which would otherwise be difficult to have as farmers, and in general, regular access to income is serving as guarantee for credit, 18 e.g. from petty commodity traders, co-workers and family members. This creates a sort of continuous debt-bondage situation (Brass, 1986), partly influenced by their dependence a non-living wage.
Again, the national labour act regulation does not address issues on delay in payment of wages-this remains a problem even for the public sector. Depending on their past employment backgrounds, some workers considered it normal, or even better than their previous workplaces. One of the farmworkers noted, 'people complain about the delay in payment because they have not worked in any other paid jobs. I worked in another private sector before, and it is better here at Volta Red. The situation is that the money is not ready. It is a company, and we have to understand 14 The minimum wage is a benchmark above which the company has sets daily piece rate, but as explained earlier, because there are no regulations to protect pieceworkers, productivity rates and output standards could be arbitrary and as such the labourers often complain about being overworked. 15 As of September 2019 16 Especially men who farm only up to one acre of tenant or own family lands 17 per calculations (from interviews) workers who had taken loans from the bank were paying interests of over 60% within 18 months 18 Credits for a wide range of things including non-self-produced items like fish, meat, oil roofing sheets, education and health needs them, they explain to us about the delay every time.' The workers' poor access to information on the existing labour laws coupled with their structural disadvantages imply that many do not even know what their rights are, and how to pursue them. One out of every three people I interviewed wanted to know about the laws on progression from casual to permanent contracts-oftentimes assuming that their employer is treating them unfairly because of their long years of work under casual contracts.
The combination of fear, misinformation, spatial remoteness, and differentiated livelihood needs affect their ability to mobilise around their working conditions (Gyapong, 2019). A farmworker narrated, 'we attempted some form of a strike when there was a delay in payment, but our supervisors advised us against it. There are also times when some people do not follow what has been planned. They go to work even when we have decided against it.' Permanent workers agitated about the lack of an enabling environment to unionise. Casual workers and those in security were also puzzled about being prevented by management to join the Ghana Agricultural Workers Union (GAWU) under the rhetoric of the 'company's internal security'. Although the law allows even casual workers to form unions, it is apparent that these provisions need to be accompanied by the necessary institutional support for marginalised groups. Officials at GAWU registered logistical and human resource constraints in their efforts to mobilise the workers. At the same time, prioritising uncommitted semi-proletariat casual workers who are geographically distant from the capital and other oil palm plantations in the south, presents practical challenges for GAWU which is under-resourced. 19 The farmworkers, therefore, resort to everyday individual forms of political reactions to deal with the unfavourable working conditions. This is often expressed through non-compliance, deception, absenteeism and continuous production on their own farms to ensure their basic food security/sovereignty and also to supplement their incomes. These actions are, however, not inherently empowering given the tradeoffs, particularly the negative implications for their upward mobility in the labour structure. In summary, the existing institutional setting is premised within a continuum of maintaining the status quo to the point where it even becomes repressive to farmworkers. Although farmworkers harness some here-and-now livelihood improvements, longterm benefits from wage work in agriculture is questionable.

Concluding reflections: rethinking labour regulations and benefits from agriculture
This paper adds new insights to land deals, provides empirical evidence on the regulatory and policy environments within which large scale agricultural investments take place, and examines the implications of labour regulations for the rural working poor-particularly wage labourers in Ghana. The governance framework for land investments and wage labour are largely characterised by absent, illusively present and repressive institutions that are influenced by both global and domestic factors. In the past few decades, land policies and redistributive reforms have come to the centre stage of development, and they ought to be accompanied by labour reforms as well. How land and labour are combined, are the fundamental distinguishing features of plantations or large-scale farms (Evans, 1995, p. 155), and this necessitates that attention is given to the labour question as much as it has been done in the context of land tenure issues. Farms are not mere units of agrarian production but embedded in power relations and social institutions that are deeply rooted in people's lives (du Toit, 1994, p. 380). Not fully exploring labour-related issues has implications for how we interpret and politically intervene in rural agrarian societies that are affected by large scale land acquisitions in an era of resource grabs.
Again, evidence from this study reinforce arguments that are critical of optimistic narratives about the employment creation prospects of large scale land investments; notably, when there is not a strong business case to demand labour. Even when jobs have been promised under 'consultative' agreements, it may be difficult to realise the expectations of decent jobs. The regulatory environment in Ghana demonstrates how and why the institutions that are expected to protect workers are either non-existent or inadequate. Particularly, the public sector and industryrelated labour policies are prevalent. They either exclude or leaves to the margins, agricultural wage labourers. Examples in Ghana include the Single Spine Salary Structure 20 applicable to civil servants, and the 2003 labour Act (651) which makes no provisions for farmworkers. Thus, several of the existing regulations tend to legitimise oppression and inequality instead of protecting marginalised groups. Ghana's labour sector requires substantial policy reforms -new laws and amendments to the 'industrial' mandates of the labour commission and also ensuring that policies are backed by extensive rural labour market research.
It is essential to have agriculture-specific legislations that deal with labour issues on both large and small-scale farms. If regulations are to be effective, the legal rights for agricultural workers need to be secure. Workers remain poorly informed about their rights. It is crucial that farmworkers' access to education, information and training on laws, rights and processes of unionisation are improved. Although companies require flexibility in labour hire, casual labour should not be the norm in land access agreements. At the same time, legal provisions for unionisation must be supported by policies and programmes that build the capacities of these "hard to organise casual workers". Organised action should not be limited to domains of conventional trade unions but with also support from NGOs, the formation of co-operatives, and other alliances and networks that are practical and relevant to their rural contexts in order facilitate leadership training. Some of these practices are leading to significant progress for undocumented, migrant and casual agricultural workers in some Western countries where social and migrant worker centres have been created by trade unions in coalition with various associations to create awareness and inform workers on issues such as occupational health, gender violence, and rights (Molina and Guardiancich, 2017). Certainly, farmworkers in Ghana are not alone in their struggles for better terms of incorporation on plantations as similar issues have been reported in other developing economies (O'Laughlin, 2017;Ouma, 2018). From a global viewpoint, farmworkers could also benefit from international governance institutions such as the International Labour Organisation (ILO) to address the huge vacuum in agricultural labour regulations.
Nevertheless, before any large-scale investment is to be justified, the 'why' question from the perspectives of landowners and those who appear to be attracted to wage labour should not be ignored. Regulatory measures that centralise information asymmetries tend to overlook the desperation which may underline people's 'willingness' to give out their lands and fall into full or semi-proletarianism (O'Laughlin, 2002). For example, many of the female farmworkers are forced into wage labour because the societal and gendered norms that shape domestic responsibilities such as the provision of food, education and health care. Closely linked, is the need to address discriminatory agricultural sector policies that leave some small-scale farmers with very few alternatives. Farmers' lack of access to inputs such as seeds, farm tools, water, fertilizers, agricultural extension services, among others, and unstable 19 Interview with a GAWU official in March 2019. Officials from GAWU also indicated that the Union has been instrumental in other well-established plantations in Ghana, and in some instances, they have even helped organised oil palm oil out-growers, but they still face difficulties in reaching out to all plantations, especially private businesses. 20 The Single Spine Salary Structure introduced in Ghana in 2010 marked a key labour reform in the recent past to address wage disparities http:// fairwages.gov.gh/wp-content/uploads/2018/09/Government-white-paper.pdf markets are key contributing factors to their choices. In Ghana, there is a considerable state provision that goes into cocoa production through subsidies 21 while other smallholder farmers in food crops such as maize and cassava are not equally supported 22 . At the same time, prices for cocoa beans are imposed, and food crop producers also do not have guaranteed price floors for their produce like those for milk and other commodities in many western countries. Farmers move in out of seasonal poverty, and desperation forces them to them make constrained choices regarding land transfers, which also inhibits their agency as farm workers on large-scale farms. As such, if there be any regulations, they have to tackle first, the initial conditions of host communities. Effective investment and labour regulations should, therefore, reflect the contested spaces of unemployment, underemployment, rural-urban inequalities, small-scale agriculture, livelihoods and power relations -within which these investments are expected to take place to 'benefit' the poor.