Abstract
Informal wage workers on smallholder tea farms make important contributions to world export of tea. The literature on global production networks only recently began to pay more detailed attention to conceptualizing the role that labour plays in such networks and has so far focused mainly on industrial settings – not smallholder farming. Similarly, the literature on sustainability standards has focused on wage-workers in large-scale workplaces such as factories or plantations. This article seeks to remedy this lack of attention and make contributions to both literatures by arguing first, that labour agency in export-oriented smallholder tea production in developing countries may not be advanced much by the sustainability certifications demanded by Western buyers and second, that labour agency can nevertheless be present at ‘the margins’ of global production networks even though informal rural wage workers are most often assumed to lack both ‘structural’ and ‘associational’ power. These arguments are made on the basis of a case study of on-farm wage labour in smallholder tea production in Kenya. The article finds labour’s bargaining power to be stronger in some locations compared to others and that differences between locations can be fruitfully explored by looking not just at workers insertion into the vertical structures of the global production network but by also including more ‘horizontal’ dimensions relating to the local context in which smallholder tea production and on-farm wage-workers are embedded.
Introduction
The global production network (GPN) for tea is characterized by a high degree of concentration in the downstream segments. This is related to the fact that tea is usually exported at a relatively early stage in the supply chain and thus that the most profitable stages (blending, packing and marketing) occur in consumer countries (Neilson and Prichard, 2009; Van der Wal, 2008). Around 90 per cent of Western tea trade is controlled by seven multinationals and at the global level, four companies (Unilever, Van Rees, James Finlay and Tata) dominate the tea trade (Stathers and Gathuthi, 2013; Van der Wal, 2008). Thus when a global giant like Lipton (the Lipton brand is owned by Unilever) pledged to source all its tea from Rainforest Alliance (RA) certified farms by 2015, it had a huge impact in terms of racheting up certification of tea producers who are mainly located in developing countries.
The production of tea is highly labour-intensive whether it takes place on tea estates or smallholder farms. In Kenya, which is the world’s leading exporter of tea and the focus of this paper, about 62% of production is carried out by over half a million smallholder tea farmers (Banerjee, 2012; Buch-Hansen, 2012). However, it is unclear from existing literature whether labour agency in export-oriented smallholder tea production in developing countries is advanced much by the sustainability certifications increasingly demanded by Western buyers.
The literature on GPNs conceptualizes how goods and services are designed, manufactured, marketed, distributed, consumed and recycled through global networks. However, it has only recently begun to pay more detailed attention to the role that labour plays in such networks and has so far focused mainly on industrial settings – not smallholder farming (e.g. Lund-Thomsen and Coe, 2015 or Selwyn, 2009). A similar lack of attention is found in the literature on sustainability standards where focus (for research as well as standards) has been on wage-workers in large-scale workplaces such as factories or plantations (e.g. Riisgaard and Hammer, 2011 or Barrientos et al., 2003).
This lack of attention to wage workers on smallholder farms is connected to a wider limitation in the literature. On-farm workers most often go unnoticed in the literature on smallholder farming where the farmer is taken as the unit of analysis and in popular discourses portrayed as the object whose situation is sought improved via for example public extension service programmes or donor-funded sustainability or market promotion projects. On-farm workers are in most cases assumed to be family labour, however, recent studies (including the one presented here) show that wage employment is more widespread than previously thought particularly in areas producing export crops (see also Cramer et al., 2014). On-farm wage labour whether casual, seasonal or regular often constitute a significant contribution to smallholder production (and production costs) and most often on-farm wage labourers find themselves in a comparatively disadvantaged economic position compared to that of smallholder farmers.
It is also generally assumed that waged labour in rural areas has little bargaining power as they mainly operate informally in conditions of widespread poverty with low levels of education and a large excess supply of labour (Cramer, 2008; Pattenden, 2016). In this paper, we wish to remedy the lack of attention and power attributed to smallholder wage workers and contribute to conceptualizing their role in global tea production networks by investigating potential sources of labour power.
GPN theory categorizes power as either corporate, institutional or collective. Thus in GPN terminology, the power of labour is conceptualized mainly in terms of the collective power resulting from workers organizations. When studying labour, the GPN approach has often been complemented by the concept of labour agency (see e.g. Alford et al., 2017; Coe and Jordhus-Lier, 2011; Cumbers et al., 2008, 2010; Lund-Thomsen and Coe, 2015). Labour agency in broad terms describe labours’ ability to act and to (intentionally or unintentionally) produce and reshape economic geographies through their actions (Herod, 2001). The concept of labour agency is particularly pertinent as it recognizes that labour is not merely factors of production, but actors that co-constitute the landscape within which GPNs operate (Coe and Jordhus-Lier, 2011; Cumbers et al., 2008; Herod, 2001). Nevertheless, we find that an analysis of the sources of labour power – or in other words, the sources from which labours ability to act derive from – is in need of further conceptualization. We, therefore, complement our analysis with Wright’s (2000) conceptualization of sources of labour power as deriving from both structural power (derived from workers position in the economic system) and associational power (derived from the formation of collective organizations of workers). Furthermore, as will be elaborated on in the following section, we expand Wright’s concept of associational power in order for it to be more applicable to a context where formal worker organizations are uncommon.
The distinction between different sources of labour power facilitates an analysis of the changing context within which workers operate whether relating to their position in the GPN or to the local socio-economic environment. The intertwining of these latter dimensions is captured in the GPN approach by its emphasis on the embedded nature of GPNs and it has been further conceptualized via the distinction between ‘vertical’ and ‘horizontal’ dimensions (Neilson and Pritchard, 2009, 2010). As argued by Riisgaard and Hammer (2011), labour agency within GPNs can only be understood in terms of the intersection of these two dimensions. So while the governance of inter-firm linkages (the vertical dimension) is crucial in structuring the terrain with which labour interacts, this is always mediated by the specific social relations of local production (the horizontal dimension). While Wright’s concept of associational power appears closely linked to the notion of collective power in the GPN approach, the concept of structural power seems not directly translatable to GPN terminology. Nevertheless and more importantly, structural power relates to both horizontal and vertical GPN dimensions and thus we feel offer a fruitful concept with which to operationalize the analysis of labour power.
In this article, we look at the labour power of on-farm wage labour in smallholder tea production in Western and Central Kenya. Our study of 115 tea pickers and 122 smallholder tea farmers in these areas reveal a perhaps surprising trend towards increasing labour power in the Central region despite a lack of collective organizing. In this paper, we explore this increase in labour power. The sample included data from farmers belonging to Fairtrade (FT) certified and non-FT certified tea factories (one certified and one non-certified factory in each region) and we were thus able to investigate whether certification to FT is linked to the documented increase in labour power and wages.
The data revealed that none of the differences found in relation to the conditions of on-farm wage labourers was related to certification status showing clearly that for FT, the smallholder farmer is the object whose situation is sought improved – not the on-farm wage labourers. We identify instead other main causes of increased labour power and earnings. Earnings are found to be related to geography (and thus indirectly, labour power including education level, labour scarcity and access to alternative labour markets) and the nature of remuneration (workers are paid less on day wages, more on piece rates). Labour power, in addition, relate to specific features of the smallholder tea product and production process. Based on these findings we argue that differences in labour power between locations can fruitfully be explored by looking not just at workers insertion into the vertical structures of the GPN but by also including more horizontal dimensions relating to the local context in which smallholder tea production and on-farm wage-workers are embedded.
This article makes several contributions to the literature on sustainability standards and the literature on labour power in general and in relation to smallholder export agriculture in particular. Theoretically, we show how Wright’s (2000) conceptualization of labour power (if modified) can also be useful outside of its original context of formal workers in industrial sectors and be applicable to contexts in which formal labour organizations are uncommon. Furthermore, we argue that this conceptualization can help cast light on the relation to labour power of both vertical (GPN-related) dimensions and more horizontal (local context) aspects – both dimensions which need to be included to adequately explore labour power amongst smallholder agricultural wage workers. Empirically, we show the importance of hired labour in smallholder export production; we explore recent developments in the power of wage labour in the Kenyan smallholder tea sector, and we investigate if these developments might be related to certification.
This article will proceed by developing the analytical framework employed to analyse the research data followed by a brief discussion of the context of smallholder tea farming in Kenya including its embeddedness in global GPNs focusing on the demand for sustainability standards. The article then outlines the methodology of the study and subsequently analyses the vertical and horizontal dimensions of labour agency along with the structural and associational bargaining power of tea pickers in Central and Western Kenya. This is followed by a discussion of labour power in small-holder tea farming in the context of sustainability standards and finally, the conclusion highlights the main findings of our analysis and their research and policy implications.
Analysing the labour power of informal wage workers on smallholder tea farms
We employ Wright’s (2000) conceptualization of sources of labour power as deriving from both structural power and associational power. This analytical framework was developed to analyse the bargaining power of formal wage workers primarily in industrial sectors (see e.g. Silver, 2003; Wright, 2000). In order to apply it to informal farm workers in smallholder agriculture, we thus need to make certain adjustments. 1
Examples of workers associational power in Wright’s terminology include unions, parties and institutional representation (Wright, 2000). However, in the context of informal workers in smallholder agriculture where formal worker organizations are uncommon, we need to expand the concept to be able to adequately analyse the sources of power. We, therefore, adopt an understanding of associational power as including all the different formal or informal types of collective efforts engaged in by workers to advance their own interests and improve their conditions. This enables us to consider even very loose and ad hoc types of organizing as a potential source of power and thus makes the concept of associational power more applicable to analyse labour power in contexts where formal organizations are uncommon.
Structural power according to Wright (2000), is derived from workers position in the economic system. Silver (2003) elaborates on this with two distinct sub-categories: marketplace bargaining power and workplace bargaining power. Here workplace bargaining power refers to the power that can be utilized from the strategic location of a certain group of workers in a key sector, for example, the ability of workers to strategically disrupt the production process (Silver, 2003). Marketplace bargaining power is a broader concept which relates to the tightness of the labour market and includes (a) the possession of scarce skills, which are in high demand by employers, (b) high employment levels and (c) the existence of alternative employment and income activities (Silver, 2003). In the context of informal farm workers, this understanding of structural power is relevant and shall be applied in the following analysis.
Keeping in mind that structural and associational power are closely interrelated, we will employ these concepts to analyse the bargaining power of tea pickers and their ability to advance their interests and improve their conditions. An important question in this regard is whether workers are able to translate structural power into associational power (Selwyn, 2012) – a question to which we shall return later in the paper.
Analysing the data from the study also revealed the importance of access to information and information sharing amongst workers as a key component of workers bargaining power. This was related to the proliferation of mobile phones and the organization of smallholder tea production in which tea pickers meet regularly at the tea collection centres. In the absence of any formal worker associations, access to – and information sharing had important implications for the bargaining power of the tea pickers and shall, therefore, be considered both in relation to associational and workplace power.
As mentioned, our analytical framework also draws on the GPN notion of embeddedness and in particular the distinction between ‘vertical’ and ‘horizontal’ dimensions (Neilson and Pritchard, 2009, 2010). The horizontal dimension refers to the local socio-economic contexts in which export industries are embedded (Neilson and Pritchard, 2010). A range of aspects might be considered here. Oseland et al. (2012) for example argue that labour agency in export aquaculture in Chile is heavily conditioned by three local and national factors (the organization of labour networks, the regulation of economic and employment affairs and the sphere of public discourse about work and its regulation). Lund-Thomsen (2013) and Carswell and De Neve (2013) in their studies of respectively the football manufacturing industry and the garment industry in South-Asia, highlight how gender, age and caste are all important shapers of labour agency, while authors such as Brammer et al. (2012), Gond et al. (2011), Riisgaard and Gibbon (2014), Lund-Thomsen and Coe (2015) in turn highlight how sustainability standards are translated into national contexts and that this translation is embedded in and influenced by amongst others business associations, nongovernmental organizations (NGOs) and workers organizations, institutional structures and the national media coverage. The concern with the local embeddedness of workers as well as our expanded understanding of associational power resonates with recent attempts in the GPN literature to expand conceptualizations of labour agency beyond collective actions by unions to incorporate more unorganized forms of labour agency (Carswell and De Neve, 2013; Coe and Hess, 2013; Lund-Thomsen and Coe, 2015). Thus, labour agency in GPNs has recently also been understood more broadly for example in terms of workers opting in and out of particular kinds of employment (Carswell and De Neve, 2013; Lund-Thomsen, 2013).
In our analysis, the vertical dimension refers mainly to the GPN-related demand for sustainability standards as well as the possible capacity of workers to strategically disrupt the production process or explore the brand sensitivity of Western companies (workplace bargaining power). For the horizontal dimension, we focus on the national- and local-level elements related to marketplace and associational power outlined above although we realize that additional factors are also likely to play a role.
The smallholder tea sector in Kenya
Tea production is an important job provider in remote rural areas where millions directly or indirectly depend on tea production and picking for their livelihoods. At independence in 1963, tea estates dominated tea output in Kenya, but current estimates show that over half a million smallholder tea farmers, who sell their tea through Kenya Tea Development Agency (KTDA) produce about 62% of Kenyan tea (Banerjee, 2012; Buch-Hansen, 2012). Tea is Kenya’s major export product and after China and India, Kenya is the world’s third largest producer of tea and the world’s leading exporter. There are currently 65 tea-processing factories owned by the 54 KTDA smallholder owned tea companies and 39 privately owned tea estate companies including multinationals, such as Unilever Tea, Finlays and Williamson Tea (Stathers and Gathuthi, 2013).
The key outlet for KTDA tea is the Mombasa auction, but there are also some direct sales. The auction system is public and prices determined by the market. Key factors affecting prices are tea supply (weather and crops), demand from export markets, movements in other world auction prices, currency fluctuations and the quality of the tea. The main buyers of Kenyan tea are Pakistan, the United Kingdom, Egypt and Afganistan, together accounting for more than 70% of Kenyan Tea export (Stathers and Gathuthi, 2013). But out of the major export markets currently, only buyers from the UK have a demand for sustainability standards. At the moment, sustainability standards in the Kenyan tea sector include the Ethical Tea Partnership, FT, RA and Utz with FT and RA having the widest spread. In 2012, 57 KTDA factories were RA certified (Stathers and Gathuthi, 2013) but currently, all KTDA factories are certified by RA, while 14 are certified by FT. The rise in RA certification has been driven by Lipton’s pledge to source all the tea for its tea bags from RA certified farms by 2015.
In terms of sustainability standards employed in agricultural GPNs however, either plantation workers or smallholder farmers are the targeted objects and as mentioned in the Introduction section, the literature on sustainability standards and smallholder production has focused on the producer or producer cooperatives – not workers on smallholder farms (International Initiative for Impact Evaluation, 2010). So far only a few studies have begun to include this group of workers (Cramer et al., 2014; Stathers and Gathuthi, 2013; Waarts et al., 2012). The findings of these studies, and how they compare with our analysis will be discussed later in this paper.
Methodology
Fieldwork was carried out in 2015 and the sampling design involved several steps. Step one was a purposive selection of four KTDA-owned tea factories. KTDA has categorized tea growing areas in Kenya into two broad geographical areas, that is, East of the Rift Valley (henceforth Central Kenya) and west of the Rift valley (henceforth Western Kenya), which differ in a number of ways. Tea acreage in Western Kenya is typically twice that of Central Kenya (Stathers and Gathuthi, 2013), while Central Kenya is much closer to the capital Nairobi. 2 Two factories from each of these broad tea growing areas were selected for this study.
Another factor that was considered was their certification to sustainability standards. For purposes of comparison, all the four selected factories were certified by RA, but only two of them were certified both by RA and FT (henceforth FT factories are marked in blue in all tables). 3
Step 2 involved selection of Leaf Collection Centres (LCCs). LCC is where farmers deliver their green leaf for weighing after which they are taken to tea-processing factories. Each KTDA factory has several LCCs. For each of the four selected tea-processing factories, two LCCs were randomly selected and from each LLC workers and farmers were interviewed. The workers were identified randomly as they plucked tea or from those who had taken tea to the LCCs. Smallholder tea farmers were identified randomly at their homesteads (see Table 1). Both workers and farmers were interviewed using a structured questionnaire although some questions were open-ended. Apart from profile information, the main themes covered included the nature of employment relations and remuneration, working conditions and efforts to improve these as well as farm management strategies. The survey data collected were processed using SPSS software and the statistical analysis is descriptive focusing on comparing the distribution of responses across groups. For reasons of transparency, we have run tests of significance (see Appendix 2) where possible, however, our argumentation builds solely on differences and patterns that could be clearly detected. 4 In addition, qualitative interviews were conducted with managers of the four KTDA factories.
Labour power
Structural power – Marketplace bargaining power
In its broadest sense, marketplace bargaining power relates to the tightness of the labour market and thus mainly concerns the horizontal context in which the GPN for tea touches down in rural Kenya and in which tea pickers are embedded. But before we dive into the marketplace bargaining power of informal tea pickers, we will briefly outline some key characteristics of our worker sample.
The 115 tea pickers (also called workers) in our sample are composed of 76 female pickers and 39 male pickers (66% females with a slight variation between regions). 5 This reflects the broader context in Kenya where the majority of tea pickers are female. The average age of the workers was 38 years old and for the majority (70.8%) of the pickers interviewed, wages from tea-plucking constituted their main source of income. Most pickers were landless, however, (24%) owned a small tea farm themselves. In the survey, we categorize the employment relation as either Casual (plucks on many different farms), Regular (hired for all tea plucking on one farm but also work on others) or Permanent (hired full time on one farm). In general, hired labour was very common with 88% of the farmers interviewed hiring workers for their tea garden (ranging between an average of 3.2 workers in peak seasons and 1 worker during the low seasons). In this sense, our study firmly does away with the long-held misconception that smallholder farmers depend almost solely on family labour.
Key indicators of marketplace bargaining power relate to whether workers possess scarce skills that are in high demand by employers, the tightness of the labour market and the existence of alternative employment and income activities. In this relation, relevant variables from our study include the level of education, worker dependence on earnings from tea picking and employer experience of labour scarcity.
For the majority (70.8%) of the workers interviewed, wages from tea-plucking constituted their main source of income and thus the tea pickers (across the two regions) can be said to have a relatively high degree of dependence on the smallholder tea farmers providing them work. However, if we look at educational levels and labour scarcity, regional differences start to appear in the marketplace bargaining power of the workers.
In our sample, the average educational level was 7.8 years of formal schooling, although with a noticeable difference in educational level between the regions, with an average of 9.6 years in Central Kenya versus 6.4 years in the Western part (see Table 2).
In terms of recruitment problems experienced by farmers, the data points towards a moderate general challenge with recruitment (indicating at least a degree of labour scarcity). Depending on location, 40–62% of farmers report facing recruitment challenges (57.4%) for Central Kenya compared to (45.3%) for Western Kenya. And although this difference is not statistically significant, 6 data from the 2009 Kenya Population and Housing Census (KIPPRA) indirectly indicate higher labour scarcity in the central region compared to the western one. Thus, a higher percentage of individuals from the central region reported having been engaged in economic activities within the last week (above the national average) compared to Western Kenya (KIPPRA, 2013). 7 Similarly both underemployment (the ratio of underemployed to total employment) and unemployment were higher in the western regions of our study compared to the central region (KIPPRA, 2013). 8 Also relevant for marketplace bargaining power is the access to alternative employment and here workers in the two areas included in Central Kenya have an advantage over those in Western Kenya in the form of stronger links to non-agricultural labour markets not least due to their proximity to the capital.
Workplace bargaining power
The other subcategory of structural power–workplace bargaining power refers to the power that can be derived from the strategic location of a certain group of workers in a certain sector (Silver, 2003). In relation to smallholder tea, it is therefore relevant to look at (a) if wage workers are able to use their insertion into a GPN and take advantage of the link to brand sensitive Western companies and Western consumers and (b) to explore what happens in case workers are able to withdraw their labour. Both issues relate to the vertical dimension of the GPN while also being highly influenced by the horizontal local context.
In terms of being targeted by the social responsibility strategies of branded companies or ethical consumer labels, informal rural tea workers working for tea smallholders are largely invisible. The fact that their labour is fragmented across time and location adds to this invisibility (see also Pattenden, 2016) and makes it unlikely that casually employed unorganized tea workers will be able to leverage consumer awareness to challenge existing working conditions. In fact, informal rural smallholder wage workers have mainly gone under the radar for NGOs promoting labour rights and/or sustainability standards as is evident in the following statement from FT: In Fairtrade we have until now focused on how to improve conditions for smallholder farmers – 80% of the men and women in Fairtrade - and on workers on plantations. Now the challenge is to unpeel the next layer of the onion and further address poverty in trade: to consider the conditions of the casual, temporary workers on smallholder farms who help out, for example, at harvest. (Lamb, 2014) The timing of the plucking is key. If it is delayed on account that you never got pluckers, you lose. If you pluck that shoot within a period of seven to nine days, your yields will increase threefold but if you now delay because you never got your pluckers and going to do so after fifteen or twenty days, your yields reduce by more than half. (9 September 2015, Area C Tea Factory Unit Manager)
Related to this structural feature of the tea crop is the way that plucking and collection of green leaf are organized. Analysing observations and interview data revealed the importance of access to information and information sharing amongst workers as a key component of workers bargaining power. This was related not to formal collective organizations but to the organization of smallholder tea production in which tea pickers meet daily at the LLCs. After plucking, each worker carries the tea leaves plucked to the nearest LLC at an appointed time. Upon weighing and registering of the tea leaves, the worker receives a receipt for the volume delivered which the worker then brings to the farmer for payment according to the agreed upon kilogram price. In the absence of any large-scale workplace or formal worker associations, this structural feature of how smallholder tea production is organized has important implications for the bargaining power of the tea pickers by making possible the access to – and sharing of information. Thus, where on-farm workers in smallholder agriculture, in general, are quite isolated, the organization of production in the tea sector where tea pickers meet daily at the LLCs provides the tea pickers with regular access to other tea pickers and access to information about tea prices and working conditions between otherwise dispersed workers. Access to and information sharing are also relevant in relation to associational power and shall, therefore, be considered further below.
Associational power
Associational power in the context of smallholder tea has so far not been leveraged by pressure from Western companies, NGOs, consumers or sustainability standards as has been the case in some plantation settings (see e.g. Riisgaard and Gibbon, 2014; Selwyn, 2007). This can mainly be explained by the invisibility of rural wage workers on smallholder farms mentioned earlier and associational power in this context thus relates to the horizontal – not the vertical context. As mentioned, we adopt an understanding of associational power that enables us to consider even very loose and ad hoc types of organizing as a potential source of power. Therefore, access to information and information sharing amongst the tea pickers as they meet daily at the LLCs can be considered as potentially yielding an (albeit softer) form of associational power because it enable workers to use that information as leverage in negotiations with farmers and potentially to coordinate demands and actions. When workers were asked ‘in which forums do you discuss tea issues with other tea pickers’ 53.6% responded that they discuss tea issues at the LLC. At the LLCs, the workers are able to share information on the different smallholder farms regarding e.g. per kilogram price paid; on-farm conditions such as treatment by farmer, access to drinking water, tea or food.
In the absence of any formal worker associations, access to – and information sharing have important implications for the bargaining power of the tea pickers and this perception is reinforced when looking at how tea workers negotiate for better terms and conditions of employment. However again a difference between geographical regions becomes clear.
As seen in Table 3, when asked ‘How do you negotiate for better terms of work?’ in Central Kenya, many workers (56.6%) responded that they discuss with other workers at the LCC and inform tea owners of the expected improvements (compared to only 6.8% in Western Kenya). This indicates a higher degree of associational power compared to Western Kenya. Furthermore, in Central Kenya, 34% of workers use withdrawal of labour (either individually 21% or collectively 13%) as a way to negotiate for better terms, while this was not perceived to be an option for workers in Western Kenya where a clear majority (88.1%) responded that they negotiated individually with the farm owner. As seen in Table 3, the patterns mentioned here are quite consistent within the two geographical regions. Thus in sum, in terms of both associational and structural power, workers in Central Kenya possess a higher degree of bargaining power compared to Western Kenya. As discussed earlier, there was a clear difference in marketplace power (stronger links to non-agricultural labour markets, higher educational level and labour scarcity) in Central Kenya compared to Western Kenya. As some sources of labour power were consistent across regions (workplace bargaining power and access to information and information sharing) it thus seems reasonable to infer that it is the stronger marketplace power which is turned into relatively higher associational power in Central Kenya – albeit not in the form of formal workers organizations. This is also manifested in the higher kilogram prices paid to workers in Central Kenya, which we will explore in the following along with the possibility that the higher degree of bargaining power might be related to FT certification.
Exploring and explaining earnings
Survey overview.
Years of schooling.
Ways of negotiation.
N=112
In Area D all workers are paid pr. day
Quantity of tea plucked per worker, kilos in high versus low season.
N = 102 out of 115 (all 13 cases missing are from area B. The cases are missing because these questions were not included until the second day of interviews in Area B).
Way of paying workers.
N=114
Worker pay per kilo.
N=83 (the 30 cases from area D are not counted here)
In Area D all workers are paid pr. day
Apart from labour power, another factor that might help explain the regional differences in kilogram prices paid to workers are differences in farmer income. While we do not have data on farmer profit, data from the four research areas for the years 2005 to 2014 show that although the kilogram price paid to farmers is the same for all KTDA factories (14 Ksh. at the time of fieldwork), the bonus that farmers receive (also known as the second payment) tend to be a bit higher for the factories in Central Kenya particularly during the last five years (see Figure 1) although the difference between the regions is only statistically significant for the years 2011, 2012 and 2013.
9
Bonus paid to farmers in the four factories over time (Ksh.).
Thus, differences in farmer income between regions might help explain some of the regional differences in kilogram prices paid to workers; however, it should be noted that the bonus is calculated on the basis of overall factory profits not farmers profit and that these might be quite different. An additional factor that might also influence the regional differences in kilogram prices paid to workers are living costs. However, although differences in living costs might have some relevance, based on available data from 2009, this was found not to offer an important explanatory variable. 10
Pay per day (highest pick × pay per day/lowest pick × pay per day).
In area D workers rely on fixed day rates, therefore the highest and lowest rates are equal.
In addition, in the areas where workers are paid per kilogram, the relatively higher daily earnings are dependant primarily on the higher amount plucked (in area C) and the higher kilogram price paid in areas A and B (see Tables 4 and 6). We thus find no indication that differences in earnings are related to FT certification. This supports our argument that earnings are related not to the vertical insertion into a standard demanding GPN but to the horizontal dimensions of geography (and thus indirectly labour power including education level, labour scarcity and links to alternative labour markets), the nature of remuneration and amount picked.
Who determines amount paid – workers’ answers.
N=115
As outlined in the section on associational power above, this perception is reinforced when looking at how tea workers negotiate for better terms and conditions of employment.
As amply illustrated throughout history, however, when labours bargaining power increases, it is always responded to by the employer side (see e.g. Silver, 2003) and the Kenyan smallholder tea sector is no exception. From the employer side, several strategies have been attempted to counter the rising labour power and related increase in labour costs (the most significant cost of production for smallholder tea growers). However, employers have so far failed to curb the increasing labour costs and increasing labour bargaining power in Central Kenya.
One strategy that has been attempted by factories in the central region has been to release a circular stipulating the per kilogram worker plucking rate to the farmers. Many farmers, however, felt the need to pay above the stipulated rate to ensure that their tea was picked in time and thus the circular had limited effect (Interview 24 April 2015, Area A Factory Unit Manager).
Another strategy – also common in labour-intensive production around the world – is the use of so-called third-party labour contractors where the responsibility for dealing and negotiating with workers is outsourced from the individual employer to a third party. Thus, KTDA has come up with what they call farm management services. The service consists in factories having a pool of pickers with a manager taking care of training and setting a fixed wage rate. This group then moves to farms who pay the KTDA factory for the service of having their farm plucked (Interview 9 April 2015, Area C Factory Unit Manager). It is unclear how widespread this practice is but it was not encountered amongst any of the respondents in our survey.
KTDA is also attempting another common employer strategy – namely to automatize labour-intensive production functions by piloting machine plucking on smallholder farms. The automatization strategy has already been implemented on some of KTDAs large-scale plantations causing quite an uproar amongst plantation pickers and their trade union (Kenya Plantation and Agricultural Workers Union). However, it remains to be seen whether machine plucking is profitable on smallholder farms due to issues of scale and quality (Interview 24 April 2015, Area A Factory Unit Manager; Interview 9 April 2015, Area C Factory Unit Manager).
Finally, uprooting or simply not maintaining the tea bushes has been known as a strategy of last resort for small farmers who deem that they cannot make their tea gardens yield profit or who choose to plant new and more profitable crops on their plots. Examples of this were encountered in the areas of investigation, but it is unclear how widespread the practice is.
Sustainability standards and wage workers in smallholder export farms
This study included data from farmers belonging to FT-certified and non-FT-certified tea factories and we were thus able to investigate whether certification to FT is linked to the increase in labour power and wages found in Central Kenya. The data revealed that none of the differences found in relation to the conditions of on-farm wage labourers was significantly related to certification status. This includes the issues presented in tables so far, but the survey also included a range of questions on labour conditions related to in-kind payments, credit and bonuses; accommodation and meals; personal protective equipment as well as the main challenges faced by workers. These results suggest, that for FT, the smallholder farmer is the object whose situation is sought improved – not the on-farm wage labourers. 13
A specific assumption has commonly been used by social sustainability standards to argue why the farmer (not the on-farm workers) is the object whose situation is sought improved via the standard initiatives, namely the long-held understanding that smallholder farmers depend almost solely on family labour and that smallholder farmers are thus not structurally dependent on wage labour (see e.g. The Better Cotton Initiative 14 ). Our study firmly does away with this assumption by showing that wage employment for export crops is much more widespread than previously thought and is thus in line with other recent studies (Cramer et al., 2014; Van der Wal, 2008).
As mentioned in the Introduction section, so far only a few studies have begun to look at wage labour on certified smallholder farms (Cramer et al., 2014; Stathers and Gathuthi, 2013; Waarts et al., 2012). Stathers and Gathuthi (2013) investigated the poverty impact of RA and FT certification among small tea producers 15 in Kenya however, the study findings are sparse in terms of impacts related to labour. In relation to per kilogram prices and daily payments to workers, interestingly, the results of the study differ considerably between producer organizations, however the report is not able to conclude on a possible relation to certification status nor to other possible explanatory variables apart from noting a general tendency towards higher per kilogram payment in Central Kenya (Ksh. 8–10/kg GL) compared to Western Kenya (Ksh. 6/kg) (Stathers and Gathuthi, 2013), a finding which is consistent with the results of our study. 16
A study by Waarts et al. (2012) looks at the impacts of RA and Farmer Field School (FFS) training in Kenya smallholder tea. 17 It finds that wages paid to workers differ (although this is not found to be related to certification or participation in FFS). The reasons given by farmers are related to the scarcity of labour (Waarts et al., 2012). This is similar to the findings of our research but again, the mentioned study does not explore the reasons for the differences beyond noting the availability of labour.
Finally, a study by Cramer et al. (2014) looked at wage labour and FT certification in smallholder and large-scale tea farms in Uganda and coffee farms in Ethiopia. Comparing across smallholder and large-scale farms, they found no evidence that FT had made a positive difference to wages and working conditions. 18 On the contrary, descriptive values and regression analysis indicated that those who worked in non-FT certified sites earned higher wages (even when controlling for size). 19 The study concludes that in the areas where the research was conducted, FT certification had failed to benefit poor wage workers. The explanation offered is that FT has overlooked their existence and that it does not rigorously monitor the wages and conditions of those working in smallholder production. Finally, certification is thought to be relatively ineffective compared to other factors that are more likely to influence both productive efficiency and working conditions. However, while the Cramer study looks at factors such as size of farm, gender and education, they do not as such explore the structural and associational dimension of labour power which we find to be such an important explanatory factor (Cramer et al., 2014).
Conclusion
Apart from showing the insignificance of FT certification to the specific conditions of workers explored in this article, our study illuminates other sources of the differences in labour power found. Employing an expanded version of Wright’s (2000) conceptualization of labour power, we have argued that labour power should not only be seen in terms of the vertical dimensions of GPN structures (such as e.g. the possible impact of ethical consumers and social standards), but also in terms of the horizontal dimension of embeddedness in local places. Employing this framework, we are able to provide important explanatory factors for the differences in earnings noted not just by our study but also by the two studies discussed above.
This article has shown how workers in Central Kenya (at least according to education level, alternative labour markets and labour scarcity) have a higher level of marketplace bargaining power compared to pickers in Western Kenya. In terms of workplace bargaining power, the prominence of casual employment relationships together with the farmer’s risk of reducing or even losing the tea yields if workers withdraw their labour even for a few days at a strategic time means that the tea pickers, in general, can be said to hold some level of workplace bargaining power. Intensifying this workplace, bargaining power is a structural feature of how smallholder tea production is organized where tea pickers meet daily at the LCCs and thus have regular access to other tea pickers and access to information about tea prices and working conditions between otherwise dispersed workers.
Adopting an understanding of associational power as including even very loose and ad hoc types of organising as a potential source of power, we are also able to understand access to – and information sharing as having potential implications for the associational bargaining power of the tea pickers. Looking more closely at this form of power (examining how better terms of work are negotiated for) however again reveals the same difference between geographical regions as were found in relation to marketplace power.
Thus in terms of both associational and structural power, workers in Central Kenya possess a higher degree of bargaining power compared to Western Kenya. As some sources of labour power were consistent across regions (workplace bargaining power and access to information and information sharing), it seems that it is the stronger marketplace power which is turned into relatively higher associational power in Central Kenya.
The associational power found in the context of informal tea workers on smallholder farms did not (at least not yet) take the form of formal workers organizations. Nevertheless, with our expanded conceptualization, we were able to consider even very loose and ad hoc types of collective efforts as a potential source of power and in the process make the concept of associational power more applicable to analyse labour power in contexts where formal organizations are uncommon.
The fact that we found a higher degree of overall labour power in Central Kenya compared to the Western part also illustrates the need to frame labour power within local and national contexts as well as within the global structures of GPNs. Thus, while the insertion into a global GPN along with some general structural features of rural labour markets and smallholder tea production might explain labour’s position in broad terms, it does not account for the variations in labour power in different locations. Differences in horizontally conditioned structural and associational power are critical for explaining these variations, and for developing pro-labour strategies that are sensitive to the specificities of the local context while not excluding the possible leverage created by linking up to broader movements.
Footnotes
Acknowledgements
We would like to thank the various workers, farmers and tea factories for their openness and collaboration during data gathering in Kenya. Valuable feedback on earlier versions of this paper was received from Peter Lund-Thomsen as well as from anonymous reviewers.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors received funding from The Consultative Committee for Development Research (FFU) 10-107DIIS.