Misconceiving 'Seasons' in Global Food Systems: The Case of the EU Seasonal Workers Directive

This article discusses the EU Seasonal Workers Directive alongside case study data of seasonal agricultural work in Spain. The conceptual contribution is to critically consider ‘seasonality’ and the related assumptions around temporary labour migration for agricultural work. This consideration informs an analysis of the Directive’s policy approach alongside its three global objectives. It is argued that this Directive is likely to fail to meet all three of these objectives; the assumed timeframe for labour demands does not correspond with unmet seasonal challenges; the lack of options for undocumented workers already in the EU may compound their marginalisation; the policy approach of circular migration and limited worker protections does not do enough to prevent new seasonal workers from falling into situations of vulnerability and undocumented status.

El Ejido (Almería, Spain). From this relatively small (and formerly poor) area comes a large percentage of a number of agricultural products retailed in all European supermarkets. The economic model of El Ejido is heavily reliant on seasonal workers. Thus, empirical research from this region provides us with key insights allowing us not only to determine the extent to which the SWD can achieve its proclaimed goals, but also what its likely implications might be.
Data from fieldwork in El Ejido in which I carried out 33 semi-structured interviews provides an overview of difficulties faced by workers there. Interviews are supported by observational field notes from my stay in El Ejido. 3 Empirical research was carried out between April and May 2012 as part of a project considering working conditions for migrants in organic and non-organic agriculture. As such, the comments by interviewees may have reduced relevance. However, as far as I am aware, many of the challenges in terms of labour relations in the context remain constant, as for example, both the financial crisis and competition from other producing regions (such as Morocco) were already underway by 2012. 4 The article is structured in three sections. A background is provided on the economic model of intensive agriculture in Section 2. I outline the key objectives and the crucial content of the Directive in Section 3, and in Section 4 the empirical data is used to examine underlying assumptions and expose some key implications of the Directive.
The article aims not only to contribute to the growing literature on seasonal workers in European law, but also to broaden the terms of the discussion of this Directive, moving beyond the legal-dogmatic analysis of positive law and towards a contextualised assessment, able to raise crucial relevant and critical questions. 5

| INTENSIVE AGRICULTURE, GLOBAL FOOD SYSTEMS, SEASONAL AGRICULTURAL WORKERS AND MIGRATION
A good deal if not most of agricultural work may well be said to be seasonal by nature. Yet seasonality in agriculture is no longer inseparably tied to the passing of the seasons of the year. Many 'seasons' have been stretched, or re-orientated alongside the development of demand-driven agri-food systems. This is fundamentally related to the way in which global food systems have transformed during ongoing processes of industrialisation and urbanisation, and more recently neoliberal globalisation. Particular food systems for particular food groups, individual food chains, and the global production networks that they involve thus function together as part of broader 'regimes' . McMichael has described these developments in terms of the development of 'food regimes' in which transforming structural power dynamics of international trade, labour and capital are engaged in the production, control, and distribution of food in particular ways that support broader political and economic structures within an overarching 'global food system '. 6 Although global food systems have altered significantly over the last decades, their development has largely been pushed forward under a so-called 'productivist paradigm' pursued initially in the 1930s, then during the post-war era by state-led investment, and more recently by private actors. 7 This paradigm has facilitated the wide-ranging 3 Participants were contacted primarily through the 'snowball' technique in which contacts with early participants lead to further contacts in the field. Some producers were also identified via open access data available via the Andalucian government. See Junta de Andalucia Sistema de información sobre la producción ecológica en Andalucía (Information system on organic production in Andalucia), at https://ws142.juntadeandalucia.es/agriculturaypesca/roae/ (accessed 31 August 2017).
intensification and then financialisation of production with public and private actors pursuing the overarching goal 'to increase outputs and the efficiencies of labour and capital for increasingly urbanised populations' . 8 We can therefore see that as production is intensified and primarily orientated towards urban consumers, the imperative for cheap food is prime and wages in agriculture are structurally restrained, 9 meaning that many jobs in agriculture have become low pay jobs. The so-called 'supermarket revolution' of recent decades has permitted demand-driven food systems to become the norm. 10 The economics of agricultural labour have therefore been transformed in recent decades and migrant workers have played an important role in this transformation. While profits are typically high for distributors and retailers (supermarkets), to a considerable extent, this is at the cost of keeping the price of products, and particularly the necessary labour costs, very cheap. 11 Agricultural workers endure arduous working conditions and flexible (and to a large extent unforeseeable) working times and working hours. In the context of increasing urbanisation and a progressive exit from farming of younger workers, migrant workers have to a large degree filled the void in those sectors of agriculture in industrialised countries where mechanisation has not taken away a need for manual labour. 12 This is particularly the case in horticulture where tasks such as fruit picking and flower cutting have remained jobs requiring a significant amount of manual labour. 13 The European Commission has indeed stated that EU economies face a 'structural need for seasonal workers' , 14 most of which would be working in the agricultural sector. The total number of seasonal workers (mostly in agriculture and horticulture) is estimated to be at the very least 100,000 people a year. 15 3 | SEASONAL WORK IN EUROPEAN LAW: THE CONTEXT

| The drivers of the European regulation of seasonal work
The European regulation of seasonal work can be said to (i) be part of the general process of Europeanisation of migration policy; and (ii) have been prompted to tackle the proclivity to hire as seasonal migrant workers in an irregular administrative situation.

| The Europeanisation of migration policy
The regulation of seasonal workers in European law is part of the protracted process of Europeanisation of migration policy. Once the Single Market was defined as an area without borders, leading in due course to the withdrawal of internal border controls, the functional autonomy of national migration policies was inherently challenged and the process of 'communitarisation' of migration control was earnestly begun with the Treaty of Amsterdam, later supported by the Lisbon Treaty, whereby internal opening was accompanied by the external strengthening of 8  borders. 16 Any third-country national entering the Union through one Member State could now move within the area without internal borders. This implied that individual Member States were able to take decisions that could deeply affect the migration policies of all other Member States.
The common migration policy has been slow in developing and has been marked by tensions regarding its ultimate objectives. This has in no small measure been influenced by the differing preferences of national governments related to the various socio-economic, demographic and economic circumstances prevailing in different Member States.
While for some time the Commission aimed to define a homogeneous status of third-country national, for whom there would eventually be a possibility to gain 'permanent residence', 17 such an approach has been largely abandoned.
Instead, during the last decade we have seen emerge a complex set of personal statuses of third country nationals, ranging from 'blue card' highly skilled third-country nationals to seasonal workers. Even a superficial comparison between the status of workers accepted under the 'Blue Card' Directive and the SWD reveals that the latter have fewer rights and more obligations than the former, particularly when it comes to the right to access and remain in different EU territories and to socio-economic rights while legally staying in them. 18

| The focus on migrant workers
Seasonal work emerged as a sector where European regulation was considered by the Commission to be necessary given the high rate of 'illegally staying workers or otherwise unregistered workers' being hired as seasonal workers. 19 Some academics suggest that until recently the employment of undocumented migrant workers in sectors such as agriculture (particularly in Southern Europe) was convenient for governments who could turn a blind eye and avoid debates on legal migration demands and figures. 20 In Spain, the period between the 1980s and the mid-2000s can be seen as having significantly altered after regularisation programmes during the 2000s in which many such migrant workers gained legal authorisation to work. 21 In recent years this situation has changed drastically, in part through the legislative changes through the EU, but also through increasing border control with countries of origin such as Morocco.
European regulation was thus justified in the name of not only 'fighting illegal immigration' , but also as an instrument to prevent the exploitation of seasonal workers. The European Union further proclaimed as a complementary goal creating the conditions for seasonal workers entering the Union legally, and as a result, enjoying fully the set of rights acknowledged to them. See Council Directive (EC) 2009/50 on the conditions of entry and residence of third-country nationals for the purposes of highly qualified employment. OJ L 155/17 (Blue Card Directive). The Blue Card workers, for example, can arrive accompanied by their families (who also have access to the labour market) and can travel freely to other Member States (see Blue Card Directives para. 14, arts 15, 19). Seasonal workers have neither of these rights (SWD art 22). Differential rights for different categories of workers does then have the effect of constructing a policy environment of indirect discrimination in which seasonal workers suffer not only from low pay but also from a second class legal status when it comes to their citizenship and labour rights in comparison to other categories of temporary workers or EU nationals.

| The regulation of seasonal work for 'development'
A high proportion of third-country seasonal workers employed in European agriculture originates from developing countries. A further aim of the European regulation of seasonal work was to 'contribute to the development of third countries' . 22 The achievement of this goal is said to be promoted by fostering seasonal migration as a form of recurrent, or circular migration. Circular migration is considered to be 'the temporary, recurrent movement of people between two or more countries mainly for purposes of work or study' . 23 This approach is sometimes seen as a 'win-win-win' for countries of host and origin as well as for migrants themselves because migrants and countries of origin may gain the benefits of migration (including remittances) while not losing long-term ties to countries of origin (avoiding brain drain). Yet many emphasise that this relatively new term (circular migration) refers to the long-established phenomenon of 'temporary migration' in which migrants are expected to arrive to host countries to work for short periods of time and then return to their home countries. 24 Regulation of seasonal work as 'circular migration' is therefore an attractive option for policy makers but susceptible to many of the same difficulties of older temporary migration programmes. These objectives can be seen to reflect the policy concerns of the EU, identified in the section above. In the following sections I will draw on the themes of these objectives and how they actually appear in the case of El Ejido.
The newly coined legal status of seasonal worker is defined by reference to the objective nature of the work ('seasonal work'), the subjective identity of the worker ('the seasonal worker') and their specific rights and obligations.

| Seasonal work
The SWD defines seasonal work as an 'activity dependent on the passing of seasons', or more precisely, an activity that is tied to a certain time of the year by a recurring event or pattern of events linked to seasonal conditions during which required labour levels are significantly above those necessary for usually ongoing operations. 26 It is left to Member States to establish which specific economic activities are to be regarded as fitting into this definition. Nevertheless, the documents that make up the legislative history of the SWD, and in particular the impact assessment, 27 make it clear that most third-country nationals who are seasonal workers are employed in the agricultural sector, with those in tourism representing a smaller fraction of the total. 28 Much seasonal work in intensive agriculture is in fact directed away from natural seasons and is re-orientated towards the market for the 22 EU Commission, above, n. 14. 23 United Nations Commission for Europe, Defining and Measuring Circular Migration (Conference of European Statisticians, 2016). 24 Castles and Miller, above, n. 20. 25 EU Commission, above, n. 14. 26 SWD, art. 3(c). 27 EU Commission, above, n. 14. 28 Ibid.
produce on the demand side. The spatial and temporal dynamics of intensive agriculture have therefore deeply changed, leading to many counter-seasonal production enclaves. 29 In these regions seasonal work therefore often takes place at precisely the times of year during which production would not traditionally have taken place (e.g. tomatoes grown in Southern Europe during the winter) and during which production in the importing countries is very difficult.

| Seasonal workers, their rights and obligations
The SWD applies only to third-country workers normally resident in non-EU countries; in contrast, the Directive does not apply to persons usually resident in the EU. 30 The SWD, as indicated in its title, sets outs the conditions of both the admission and stay of non-EU citizens (third-country nationals) entering the EU in order to be employed as seasonal workers. 31 In order to be admitted to work in the EU under the SWD, workers must accompany their application with a binding job offer or a work contract that states details such as the place and type of work, pay and paid leave and working hours. 32 Evidence of adequate accommodation is also required. 33 Seasonal workers are guaranteed equal treatment with national workers of the host country, 34 extending to terms of employment including minimum working age, working conditions, pay and dismissal, leave and holidays and health and safety requirements at work. 35 It also includes the right to freedom of association including the right to strike in accordance with national law and practice. 36 It applies to some areas of social security. Member States must also establish a complaints mechanism in order for seasonal workers to be able to complain against employers. 37 The SWD requires EU Member States to introduce measures to prevent abuse, and sanctions for any abuses that might occur. 38 However, EU Member States may restrict equal treatment in some areas such as for unemployment and family benefits, education and vocational training. 39 In particular, seasonal workers do not have any route towards a permanent right to stay in the European Union. The SWD establishes that a maximum stay for seasonal work will be between five and nine months in any 12-month period, and that this maximum must be determined by each Member State. 40 Seasonal workers are permitted to extend their stay once within this time period to extend their work contract, or change their employer, if they meet the conditions for entry and if there are no grounds for refusal. 41 If still within the maximum time period permitted to stay, EU Member States may permit workers to prolong their contract with the same employer and extend their stay more than once. 42 In terms of circular migration, the SWD provides for the facilitation of the re-entry for seasonal workers who were previously admitted to the same Member State and who respected the relevant conditions during previous stays. 43 29 Gertel and Sippel, above, n. 1. 30

| EL EJIDO AS TEST CASE OF THE EUROPEAN STATUS OF 'SEASONAL WORKER'
El Ejido can be considered a very fitting case through which to consider the way the SWD responds to some of the current challenges related to seasonal work. This Southern Spanish region is a microcosm through which to study the potential impact of the SWD, in particular, the extent to which it can be a means of grappling with the different objectives and interests at stake.

| El Ejido as 'a global enclave': Economic model and social realities
When the local social and agricultural organisation and the economic, political and even legal institutions of a place appear to be functioning together in a way that facilitates the establishment of intense globally orientated production, specialist scholars use the concept of 'global enclave'. 44 The characterisation of El Ejido as such a 'global enclave of production' is useful to capture the sense in which El Ejido represents an acute but not atypical case of intense export-orientated food production in Europe. This is also a case where workers are very vulnerable to both global and local changes in the political economy, and in related labour arrangements, in part because it is so involved in the dynamics of globalisation and global trade.
The production model is relatively straightforward. Production is concentrated on a standard group of horticultural crops (tomatoes, peppers, aubergines, courgettes, cucumbers, green beans, melons and watermelons). These are predominantly cultivated by small farmers, counting on the work of migrant workers. Farmers sell produce either to 'cooperatives' that act as intermediaries and pack-houses or to auction-houses where buyers attend in person to buy loose, unpackaged crops.
Jiménez Díaz estimates that the 30,000 hectares of greenhouses in the El Ejido area are worked by 20,000 small-scale farmers (owners of the 1 hectare and larger plots) and 40,000 migrant workers. 45 This suggests that a significant number of migrant workers live in El Ejido without residency permits. It is unclear how many undocumented migrant workers are living and working in El Ejido. While officials assert that irregular workers account for only a 'small minority' (indicating that this number would be many less than those 26,000 with residency permits), 46 local union members estimate that the number could be much larger. 47 Only a small number of migrant workers enter the workforce in El Ejido through the kind of programmes of 'recruitment in countries of origin' that the SWD has set out to regulate and unify across the EU. 48 However, one aim of the SWD appears to be to promote the option in regions such as this of hiring workers who come on fixed-term visas. Such programmes would therefore fulfil the objective of closing another door to irregular migrants and facilitating migrant workers to live inside the EU when, and only when, they are needed by employers.
In recent years the case of migrant workers in El Ejido has drawn attention from international organisations, including the International Labour Organisation (ILO). In several cases the ILO has investigated issues which are directly relevant to this research. Two direct requests adopted by the International Labour Conference both requested 44 C. de Castro, E. Gadea and A. Pedreño, 'Inmigración, crisis del sur de Europa y sostenibilidad social de las estrategias de desarrollo: El caso de los enclaves productivos de agricultura intensiva' (Immigration, Crisis in the South of Europe and social sustainability of development strategies: The case of the enclaves of agricultural production) (2014) 11 Revista Trabajo, 89. demonstrated concern about overall labour market conditions in light of the recent labour reform of 2012, 51 and in a committee report of 2014 the ILO also requested to be kept up-to-date on legislation limiting collective bargaining. 52 Although the foundation for labour rights in Spain is strong (Spain has ratified more ILO conventions than any other country), 53 the particularly vulnerable position of seasonal migrant workers has drawn concern from both local and international forums. Of particular relevance is that the ILO has expressed concern about 'the situation of immigrants in the context of the economic and financial crisis' and specifically requested information about 'the situation of migrant workers in El Ejido' . 54 This ILO concern is supported by socio-economic research in Spain which has shown that an especially high burden of the social effects of the economic crisis is carried by the foreign resident population who are particularly likely to have fallen into poverty and unemployment. 55

| The impact of the seasonal workers directive in El Ejido
The extent to which the status of 'seasonal worker' as defined in the Directive can actually contribute to the achievement of the self-proclaimed goals of European regulation can be open to serious question on the basis of the economic and social realities of intensive agriculture in the European Union, as the case study of El Ejido shows.
In particular, the Directive characterises seasonality in ways that do not reflect the actual seasonal challenges faced by farmers either (i) because labour demand is not seasonal at all (because crops are produced almost throughout the year), or (ii) because seasonal demand is ultra-short, a matter of days, not weeks, less so months. Moreover, circular migration, far from contributing to development, weakens the status of the seasonal worker by neither providing clear gains to their countries of origin nor offering them any long-term prospects in their host countries.
Finally, the new status is one extremely difficult to acquire for those seasonal workers who were already working in Europe, at the very same time that the lack of fit of the Directive with economic reality makes it probable that farmers will prefer to continue hiring such irregular workers to some extent; the net result is likely not to extend the legal rights to these seasonal workers, but to condemn them to continued illegality, and further exposure to exploitation.

| Seasonal no more? A season of ten months
If we move from general assumptions to actual practice, the seasonal story is more complex. Industrialised agriculture, and particularly in cases of 'global enclaves' such as El Ejido, has been almost de-linked from ecological seasons.
Technological innovations in the 1960s and 1970s involved drilling into aquifers for water, the construction of greenhouses, planting in sand, and agro-chemical inputs that meant farmers could achieve the necessary climatic such as tomatoes and peppers would have occurred. In as much as seasonality is still relevant in El Ejido, this refers to a 'counter-season' occurring when the traditional season would not have been. This counter-season could be said to still be tied to a certain time of year but this time of year is as much defined by the market as by ecological seasons.
The season in these terms in El Ejido refers to the time of year when crops such as tomatoes are costly to grow in Northern Europe and therefore command a high enough price for farmers to produce and sell them in Southern Europe.
Besides revealing a certain detachment from agricultural economics on the side of the drafters of the Directive, the 'contextualised' definition of (un)seasonal agriculture implies that some of the normative solutions, which may be regarded as coherent in a context of 'classical' seasonality, are set to be highly problematic. A 10-month season cannot reasonably be understood to represent a period of time during which 'labour levels are significantly above those necessary for usually ongoing operations'. Rather, the labour demand would be more accurately characterised as 'ongoing' or 'constant'. By the same token, where farmers do have unmet needs for workers with permission to work and reside inside the EU, it would be just as likely to be for a long-term and ongoing role, as for the 3-9 month periods outlined in the SWD. Of course, many farmers may have shorter planting periods or may require workers for shorter periods of time and I will explore one instance of this. However, it is clear that there is an ongoing demand for seasonal workers for at least 10 months of the year. Consequently, it seems reasonable to question whether the labour market for almost year-round workers is best met by a temporary workforce who are not officially based in Spain.
Moreover, it is the 'false' seasonality of agricultural work that goes a long way towards explaining the reluctance of farmers to hire workers through circular schemes as that foreseen in the SWD. The new 'official' procedure is restrictive, and may end up encouraging farmers to work with undocumented workers in open-ended, longer-term working relationships despite the potential penalties that they may face. The small size of many farms in El Ejido leads quite frequently to farmers working alongside one or two workers on a daily basis and developing deep personal relationships with them (which both parties have a clear interest in developing). We also now have to add a legal framework that is ill suited to the actual economics of agricultural production (and impinges upon the personal relationship becoming an enduring one, given the temporal limits on their stay on European soil). Certainly historical practices of what Gonzalez-Enriquez referred to as 'the Cheap Model' also have their bearing and influence. 57 Until quite recently, the norm for farmers was to hire first and then rely on regularisation programmes to work with migrants to eventually regularise their status. But the issue is not merely the legacy of a problematic set of practices, but is, as has just been stated, one closely related to the economic realities and the societal ties of a given mode of production. 'Irregular' or 'informal' hiring is not an option of choice for most farmers. Still, the fact that some (if not many) farmers are still willing to hire undocumented migrants for ongoing, rather than just for casual work, demonstrates the long-term nature of the working relationships that farmers aim to develop with workers, as well as the investments they may be willing to make for workers whom they work well with. In these kinds of working relationships, farmers are looking for reliability and medium-to long-term availability, something that a 3-, 6-or 9-month work visa, with no guarantee of year-on-year renewal, may not correspond to. For small farmers, hiring workers in their countries of origin may not therefore meet this need for workers that can be hired quickly and personally, with long-term permission to stay and who can be seen as potentially long-term employees, available to work for the majority of the year.

| The watermelon dilemma: The ultra-short demand for seasonal work
The section above argues that there is a near year-round demand for 'seasonal' workers in El Ejido due to long seasons which often comprise the majority of the months of the year. There is an ongoing need for workers in planting, 56  pruning, spraying and harvesting labour-intensive crops such as cherry tomatoes and peppers. However, it is still the case that the production pattern of other crops results in sudden peaks of labour needs due to certain crop cycles.
This situation is most clearly illustrated by the case of watermelons. Watermelons, unlike cherry tomatoes, or peppers, require little year round labour, yet, at the moment in which they are ready to be harvested, they require many more workers than would be needed for their cultivation at other times. In such cases, there is clearly a 'seasonal need' for a larger labour force, but for periods much shorter than those envisaged in European law: a day or two at a time at the very peak of harvesting or planting.
In my interviews with farmers, four methods of hiring in peak times emerged: relying on (extended) family, requesting the help of employees of fellow farmers, attempting to hire temporary workers through formal channels, and hiring workers waiting at roadsides.
Firstly, some farmers emphasised the role of unpaid family labour and the tradition of bringing family members together at harvest time to help out. Secondly, farmers told me about their personal strategies to balance legality with peak needs for workers by 'swapping workers' . In this scenario farmers work with networks of family and friends and ask their employees to work on a different farm for several days, fellow farmers then respond by asking their own employees to go and assist on the other farm when the time comes for the harvesting of the corresponding crop.
In this second scenario farmers managed to avoid the strictly illegal practice of employing someone without a contract and social security with the a-legal practice of asking their employee to work on another farmer's plot. Thirdly, some temporary agencies existed, which particularly larger employers used. However, for some smaller farmers the time and costs involved appeared prohibitive.
Fourthly and finally, it became apparent to me through observational research that some farmers continued to hire (documented and undocumented) workers informally for short periods without contracts. In this case farmers would balance the risk of impromptu visits from labour inspectors during this short period of time with the feasibility, hassle and cost of hiring workers legally with little notice. The data regarding this practice emerged in my inductive analysis as the narrative of the 'roundabout' . Residents in El Ejido would often discuss irregular workers at the roundabout, whether or not the spots in which they waited to be hired were indeed roundabouts or other public places such as roadsides or village squares. Significantly, one farmer also originally told me that he had no need to hire extra workers at peak times due to what he asserted as the 'size of his plot' , which was properly organised so as to be workable just by himself, his family and the one worker that he had on a fixed contract. However, on following visits, by chance I met the worker on the fixed contract who was clearly working with another man who was employed casually without a contract, and both the farmer and the worker then discussed their frustration with the difficulty of managing labour demands. The contracted worker explained how he would often go and hire someone extra to help without bothering with a contract unless it would be for several months.
The level of nervousness in the region about undocumented workers and their relation to the agricultural sector was reflected in the research as a whole and clearly showed up in data analysis. Despite the fact that interview questions were focused on work in agriculture, the theme around which most quotes were identified was that of 'personas sin papeles' or, 'people without papers', which is how both farmers and workers most often referred to undocumented workers.
Despite the attempt by some to portray the very poor living conditions of some migrants as only minor exceptions to a production model with a well-regulated labour market, and really very little to do with the agricultural sector, one example indicated the existence of a persistent structural gap in the labour market. This gap can be seen as an ongoing demand for workers who could expect to be employed at a moment's notice and paid cash-in-hand despite not having the 'papers' , for very short-term work. This, which I describe here as the 'case of the watermelon' , also applies to other crops and implies a sudden need for workers who wouldn't be needed during the rest of the season.
Once they were going to come [the export cooperative], we were arguing about the watermelons, whether we should pick them or not and in the end they said to me; "alright, this afternoon we pick them" and I said, "this afternoon I don't have any workers" and they said; "either we come and get them this afternoon or we don't come". So I said, "Well, come" and when they come to pick the watermelons, I need a lot of workers, four people, and so I say, "where am I going to find those people? I can't get…" so I went to the garage in El Ejido and there were four Africans just there. (Small farmer) In the example above, the farmer resented having to resort to employing workers at a moment's notice from the informal sector in order to comply with buyer demands. Yet, due to both the nature of the growing cycle and of the pressure from buyers, the farmer was put in a situation in which the immediate peak need for workers was not met by the workforce usually working with him. This unresolved question of very short-term seasonal work has clearly been recognised by trade unions in the region and others who have tried to develop formal ways in which to deal with it.
However, such initiatives seem to have been only partially successful and other practices for dealing with short peaks appear to be more common.
The workers' organisations here have tried to form a type of bank of workers where, for example, for watermelons it's only a day but that day I need eight people. Well, so that there's a place that you can go and get workers that are legal so that that day you don't have problems. Because if not, what? I go and harvest the watermelon, before what was done, was you told your neighbours, and all the neighbours came and helped you and when they needed to harvest you went to their plot. They called it "tornateón". Tornateón means to say that I go with you today and tomorrow you come with me, that's how it used to be done. But now things have changed, what's done is, we'll ask for the workers of a friend or if not, go to the stop and you'll see many immigrants, but those immigrants, well the problem is that they don't have papers. So there, yes, there is problem with that [seasonal work], a gap.
(Farmer who also worked as a Trade Union representative)

| Circular migration: Does it really work?
The SWD was expressly intended to promote 'circular migration' . Yet, in this context it is pertinent to ask whether circular migration as a policy objective can be realised while upholding the rights of seasonal workers. 4.3.1 | Does it make sense to foster circular migration for 'unskilled work'?
Although the ILO has come out in favour of the circular migration of skilled workers, 58 for many reasons, including the prevention of 'brain drain' and the potential of significant remittances, the case for circular migration of unskilled workers is less clear. Many have suggested that circular migration is just a new way of classifying unskilled workers as temporary and therefore depriving them of opportunities to gain long-term residency or citizenship should their work take on a long-term character. 59 It is yet to be seen whether this attempt to manage 'legal migration' and to keep it temporary and on the borders of Europe will work, as the ILO warns: Some argue that it is acceptable. Political philosophers Ottonelli and Torresi argue temporary migration defies the requirements for fixed citizenship rights and that temporary migrants have particular needs in their host countries. 62 They therefore advocate the development of a set of particular rights for temporary workers, allowing them to have, rather than secondary status in host societies, specific rights as tourists or diplomats do. 63 This theoretical position is one that could be seen reflected in the SWD, which attributes the rights to third-country workers that the European Union decision-making bodies have deemed appropriate for their particular situation. Yet, the position appears to rest on similar assumptions of both the desirability and feasibility of a temporary workforce. As discussed above, the attitude of exceptionalism to seasonal workers conceiving them as 'necessary' but only 'temporarily' needed is not well founded when we consider carefully the ongoing demand for low-waged labour. Rather than needing a different set of rights, this appears to be yet another case of an attempt to justify measures which are designed to keep low-wage migrant workers as only temporary members of the EU. Such programmes have historically invariably led to settlement, despite the many efforts to keep temporary migration from developing into permanent settlement. 64 There is no reason to expect that the case of seasonal work is any different and leaving workers less protected is simply likely to make them more vulnerable to exploitation in the process.
The vulnerability of migrant workers does appear to have been taken into account in later stages of the development of the SWD. According to the account of its development by Fudge and Olsson, a main achievement within the SWD appears to have been the securing of the right to equal treatment principle which sets out the principle of equal treatment between seasonal migrant workers and EU workers. 65 This point was also welcomed by NGOs. 66 Yet the SWD's labour rights protections are not extensive. Migrant workers are particularly vulnerable to exploitation when their legal residence status is tied to their employment with one employer. 67 This risk is therefore highly relevant in the case of this Directive, which is 'demand-driven' and where entry permits are dependent on workers having a contract with a future employer. 68 There is a provision in the SWD which aims to weaken this link, making it possible for workers to apply to extend their stay in order to work for a different employer. 69 However, in not allowing workers to remain in the country when unemployed, it is not clear how a worker facing exploitation would be able to find another employer. Underlining this danger of exploitation of vulnerable workers is what Rijken has highlighted as a 'glaring absence' of any reference to labour exploitation or trafficking, the knock on effect of this is that the Directive may fail to ensure protections are available for workers in cases of exploitation. 70 Seasonal workers would have only limited employment options as it is a requirement that their passports indicate that they are clearly within the Member State for the purpose of seasonal work. 71 Furthermore, seasonal workers' access to social assistance programmes could be limited, as they are excluded from unemployment and some education programmes. 72 Overall, this approach is very much one of looking for a special, or reduced set of rights 62 V. Ottonelli and T. Torresi, 'Inclusivist Egalitarian Liberalism and Temporary Migration: A Dilemma', (2012) 20 Journal of Political for temporary migrant workers, and therefore one that, despite equal treatment provisions, does not place them in a position where they are protected to an equal degree as EU country nationals.

| Circular migration of seasonal workers and development
In light of these limits to labour rights we might ask what, if anything, the SWD offers in terms of development?
The SWD fits into the triple-win rhetoric of the 'migration for development' logic that the EU has adopted in recent years. 73 The logic in this regard is that circular migration benefits the country of origin, the country of destination and the workers themselves. Yet these benefits cannot be understood as equally balanced due to the fact that it is the EU Member State that establishes the migration dynamic, and being demand-driven this is established to respond to the needs of the countries of destination, not primarily to the needs of either countries of origin or the needs of the seasonal workers themselves. 74 The vision of 'development' that the SWD aims to offer countries of origin is presumably based on the notion that seasonal migrants will return remittances and/or savings to their home countries as well as potentially other skills. However, while the wage differential between Spain and Morocco, for example, is significant, González Enríquez and Reynés Ramón suggest that the differential is not sufficient to be able to contribute any significant changes to living standards and nor are the numbers of workers admitted (which is left to the discretion of Member States) significant enough to have a developmental impact, for example via remittances. 75 At least 20 Member States already had specialised and widely diverging admission schemes for seasonal workers prior to the passing of the SWD. 76 Spain already makes use of programmes such as that foreseen by the SWD to hire seasonal agricultural workers in their countries of origin as noted above. As charted by the study by González Enríquez and Reynés Ramón, the programme undertook a rapid turnaround from 2004 to 2005 going from one extreme to the other. In 2004, 60% of those who entered Spain under the programme overstayed their visas partially due to corruption in the selection process. 77 Following this experience the selection process was revised and those hiring the workers established that in order to ensure that workers would not overstay their visas they should be mothers of dependent children in their home state. 78 This programme of targeting mothers to work away from their dependent children highlights the extent to which Member States' individual interpretations of circular migration seasonal work programmes have significant specific ethical problems of their own, something not foreseen by the SWD.
The ethical questions raised here in the context of examining the SWD itself therefore appear likely to multiply in relation to the specific ways in which each Member State goes about transposing this Directive into national law. This is related to the observation of Olsson and Fudge who highlight how many of the clauses in the SWD regarding labour rights use the term 'will' rather than the stronger term 'shall' , indicating that some of the provisions in the SWD may be weakened when it comes to being applied in national law. 79 In this context, the ongoing attention to how such programmes progress and how the SWD is transposed in the coming months and years is vital.

| Workers already (stuck) in Spain
The endeavour to formalise work, particularly through labour inspection, is aimed at preventing a race to the bottom by striving to ensure that employers provide workers with contracts and that all work is done with social security protections. However, taken together, the measures have also created an increasingly hostile environment for those migrants who have not managed to gain documentation in order to work legally in the region.
As one unemployed undocumented worker who had been in Spain for 3 years told me, 'Me, I'm not legal, I'm still not legal. People who have papers work with their papers.' He also emphasised that this left him and fellow workers more vulnerable to poorer working conditions, for example by not being given proper protective equipment when using chemicals in the greenhouses and feeling less able to complain.
As Olsson charts, the concern for providing an option for migrant workers already in seasonal agricultural work was raised during the drafting of the SWD. 80 Stakeholders put forward proposals to allow for such undocumented workers residing in the EU to be able to apply for a Seasonal Workers Permit for a transitional period of two years in order to legalise their status, but this was not accepted. This possibility might have done something to bring the SWD more into line with the needs outlined by both farmers and workers who are seeking to legalise their own status or that of their workers. This omission was also criticised by anti-racism, migrant-rights and homelessness NGOs who argue that an option should have been included to allow third-country nationals already residing in Member States to apply to work as seasonal workers. 81 As Fudge and Olsson discuss, the EU chose in the end to focus on 'sticks' to deter more migrants from arriving in hope of work. 82 This decision demonstrates a lack of sensitivity to the vulnerable and increasingly marginalised situation of undocumented migrants already in places like El Ejido, who, as discussed above, still fill a gap (as in the 'watermelon dilemma') in the labour market which means that they cannot be detached from the seasonal labour demands of the agricultural sector, and the SWD is unlikely to remedy this.
From the perspective of a case such as El Ejido, the use of 'sticks' can be seen as disproportionate. Following the economic crisis in Southern Europe, fewer migrants appear to have been arriving in search of work in this region in any case. This is shown in local government figures where immigration can be seen to be slowing from around 2008. 83 Undocumented migrants spoke of the sensation of being trapped, having missed the era of mass regularisations as well as the periods of economic growth in the region, and the more relaxed attitude to hiring those without documents in previous decades. This can be related to multiple factors at least including: increased fear amongst farmers due to labour inspection fines; large numbers of migrant workers now with documentation; the economic crisis which has led more Spaniards back to agriculture; and finally, increased abilities for farmers to hire overseas, through policies such as those facilitated by the SWD. The need for more 'sticks' in order to discourage further irregular arrivals is therefore highly questionable. Rather, a route to legality, even if temporary, may have been an option that some vulnerable workers would have welcomed, alongside their employers.

| CONCLUSION
In this article I have considered the Seasonal Workers Directive in the light of experiences of seasonal workers and their employers gathered through empirical research in the global enclave of El Ejido. I have shown that there is a clear mismatch between the proclaimed objectives of the Directive and how it works on the ground. Firstly, the Directive defines seasonal work in ways that do not correspond to the economics of food production in the areas where demand for seasonal work is most intense. Secondly, the proclaimed aim of fostering 'circular migration' as a means of development does not correspond with the low developmental potential of the irregular, insecure and poorly paid work that is offered to seasonal workers. Thirdly, no specific provisions are included to respond to the situation of workers who are already involved in irregular employment. 80 Olsson, above, n. 17. The main assumptions dealt with throughout this article have been around the nature of seasonal work and the nature of the labour demand that this creates. Through the examination of some of the actual challenging labour dynamics faced by farmers and workers, I have found that the Directive is ill-fitted to the economics of agricultural production and creates a serious risk that the 'guarantees' provided to workers may remain dead letter. Meanwhile, labour needs may continue to be met in informal when not plainly irregular ways, thus fostering rather than an end to irregular work, a weakening of the position of seasonal workers. The assumed nature of seasonality is met with a programme of work permits for periods of several months. Yet 'seasonality' , rather than responding to a particular time period of unmet labour demands, appears to function more as a legitimising factor for a new system of temporary labour migration which aims to control unauthorised migration whilst ensuring sectors such as horticulture continue to attract low-paid migrant workers.
Furthermore, the Directive, by introducing a unified way through which seasonal workers for this sector can enter the EU on unequal terms to those entering for other purposes (such as those entering via the Blue Card Directive) reinforces the structural status of agriculture as a second-class sector to be manned by second-class workers, offered fewer rights and options than those entering the EU to work in other sectors. In this sense, the chosen case study of El Ejido (situated in Almeria, Southern Spain) is rather typical, if not paradigmatic, of intensive agricultural export-orientated production in its dependence on poorly paid migrant workers.
In aiming to construct a Directive which would not challenge the status quo of the sectors involved, nor of the current model of production and distribution, but rather meet the labour demands of the current system with those vulnerable due to their temporary status, this Directive appears to do more for employers than for workers. Moreover, it reinforces the necessary effect that agricultural work will be low-paid and remain a sector for those workers who are most vulnerable in society.
Perhaps most seriously, the SWD meets the challenges of undocumented migrants who already carry out seasonal work in agriculture with no options. This leaves undocumented migrant workers in Europe no further forward and potentially further marginalised as their employers have more options to contract workers from outside of Europe. While this approach could be interpreted as firm discouragement for new undocumented workers to arrive in search of work, it is not clear that it will do any such thing. The scenario that I have discussed as 'the watermelon dilemma' is likely to persist; therefore, the extremely casual and unattractive short-term work is likely to continue to be open to undocumented workers.
Finally, the protections available to new workers entering Europe under this Directive are insufficient and leave them vulnerable to falling into the same pool of undocumented workers resident in the region, looking for evermore scarce employment. In light of this, and the potential meagre prospects for development via circular migration for seasonal work, we can see the SWD as failing to live up to its objectives.
At this stage it is uncertain whether the situation as described in this article will change in the coming months and years due to the impact of the SWD or whether this will simply consolidate current practice and further promote already existing Spanish circular migration programmes. In either case, the SWD does indicate a policy approach to seasonal work and circular migration over which Member States (except the UK, Denmark and Ireland) have come to agreement and could therefore signify the current direction in which EU approaches to seasonal work and similar low paid work may take in coming years. This policy approach outlines a concerning differential approach to the rights of temporary workers in relation to EU workers. Social actors should therefore remain as alert as ever to the ongoing need for further efforts to protect migrant workers, particularly those undocumented workers outside the scope of this Directive.