Fictitious commodification and agrarian change: Indigenous peoples and land markets in Highland Ecuador

Creating private property rights and establishing land markets were fundamental to the historical development of capitalism in the Global North and remain at the centre of capitalist development in the Global South. This article contributes to debates about these processes by analysing the relationship between land markets and indigenous peoples in Highland Ecuador. Building on Karl Polanyi's concept of fictitious commodities, it elaborates a new concept that makes an analytical distinction between the activation and development of land markets. The former refers to the occasional participation of actors in markets to secure land, whereas the latter relates to the establishment and expansion of markets that regulate the distribution and value of land through market prices. Focusing on indigenous land struggles in the late 20th century, this article shows that the activation of land markets created opportunities for indigenous peasants to secure land, whereas the development of land markets closed them down. Social and class differentiation among the highland indigenous population increased through this contradictory process. The article connects this historical analysis to recent developments in Ecuador to contribute to empirical and theoretical debates about contemporary land struggles and agrarian change elsewhere in the Global South. ‘ To isolate [land] and form a market for it was perhaps the weirdest of all the undertakings of our ancestors ’

valuation and exchanging it at market prices, and mestizo and, especially, indigenous peasants tending to assign alternative meanings and values to land and largely circulating it through family and community institutions. Thus, in the late 20th century, land markets coexisted with alternative forms of exchange, which were underpinned by different practices and relations (Goodwin, 2017).
In this article, I explore this historical process by analysing indigenous efforts to secure land through the market during land reform in the late 20th century. 2 I elaborate a new concept to understand this process which makes an analytical distinction between two dimensions of the land market: activation and development. The former refers to the occasional participation of actors in markets to secure land. The latter relates to the construction and expansion of markets that regulate the distribution of land through market prices and enable and promote the economic valorization of land. I show that these two processes occurred simultaneously under land reform, with activation creating spaces for indigenous peasants to obtain land and development closing them down.
In developing the activation-development concept, I take inspiration from Karl Polanyi, especially his claim that land is a 'fictitious commodity' that cannot be reduced to a factor of production or item of exchange (Polanyi, 1944(Polanyi, /2001. 3 Land, as Polanyi urges, 'invests man's life with stability; it is the site of his habitation; it is a condition of his physical safety; it is the landscape and the seasons' (Polanyi, 1944(Polanyi, /2001. Hence, it has important non-economic dimensions, including social, political, spiritual and ecological. This suggests that even if individuals or families purchase land on the market and secure private property rights, they will not necessarily treat it as a commodity and subject it to strict economic valuation. 4 I will argue that this was generally the case for highland indigenous peasants in Ecuador in the late 20th century. The private land they secured through the activation of the land market was wrapped up in a long-term social and political struggle and performed multiple non-economic functions. Thus, it was generally not reducible to a commodity freely available for exchange on the market. Rather, indigenous peasants tended to circulate private land through inheritance and marriage (e.g., the formal transfer of land to children following the death of the head of the household) and buying and selling between family and community members (e.g., the sale of land to a member of the same community). The exchange of this land was generally, if not universally, regulated through communal authorities and relations, which were underpinned by organizational structures indigenous communities developed around the Ley de Comunas (Registro Oficial [RO] 558,1937). This law was originally introduced in the 1930s to govern indigenous peoples; however, it provided indigenous communities with a tool to strengthen communal organization and challenge the elite and the state in the late 20th century.
Establishing legally recognized communities, which was the dominant but not universal pattern in the highlands, was part of a broader process of collective organization that helped activate land markets. Collective mobilization prized open markets and created space for indigenous peoples to obtain land. Protests, occupations and strikes were among the visible tactics indigenous peasants employed to encourage elites to sell them land. Collective mobilization was also used to pressure the state to increase the regulation of land markets through land reform (Goodwin, 2017).
Hence, while indigenous peasants were tapping into markets to obtain land at the local level, they were mobilizing to limit the reach and influence of land markets at the national level. I argue that these two features of highland indigenous mobilization were part of a 'countermovement' that aimed to decommodify land by regulating and delimiting land markets and protecting and expanding alternative forms of organization and exchange (Goodwin, 2017(Goodwin, , 2018Polanyi, 1944Polanyi, /2001. However, indigenous engagement in land markets in the late 20th century increased social and class differentiation within the highland indigenous population, which ultimately weakened the countermovement and decreased the possibility of broad-based land mobilizations occurring in the early 21st century (Bretón, 2015a(Bretón, , 2015bGoodwin, 2017;Henderson, 2017Henderson, , 2018. Thus, I do not offer a static-or essentializedinterpretation of indigenous resistance but argue indigenous peasants and movements evolved as market capitalism advanced. 2 For details on the size, distribution and evolution of the highland indigenous population in the late 20th century, see Zamosc (1995). 3 Polanyi's definition of 'land' is broad, including various elements of nature. In this essay, I adopt a narrower reading, focusing on the integration of land into markets. More broadly, I follow the exposition of Polanyian social theory I outline elsewhere (Goodwin, 2018). 4 Numerous scholars have recognized this point in relation to rural land and property in other contexts. See, for example, Ferguson (1985), Gudeman (2001) and Li (2014).
In making this argument, I both support and challenge Polanyian social theory. 5 On the one hand, I show that land markets are shaped by social and political as well as economic forces (Topik, 1999). The land market that emerged in the Ecuadorian Andes in the late 20th century was not the atomized market imagined by neoclassical economic theory. Sustained state intervention was required to create and maintain the market and social and political factors influenced both the supply of and demand for land. For example, the decision of elites to bring land on to the market was strongly influenced by indigenous mobilization and was therefore not simply a response to changes in market prices and conditions. I also show that land has multiple non-economic dimensions, highlighting the relevance of the fictitious commodities concept and cautioning against reducing land conflicts to purely material struggles. Furthermore, I demonstrate that markets exist alongside other spheres of exchange in capitalist societies, with mechanisms based on alternative relations and practices coexisting with land markets in the rural highlands (Goodwin, 2017;Polanyi, 1957). On the other hand, I argue that when combined with collective mobilization and state intervention, land markets created opportunities for indigenous peasants to obtain land and, in some cases, reverse historical processes of colonization and dispossession. I also show that land commodification sets in motion changes that can limit resistance to the market over the long run, highlighting the need to historize and contextualize land commodification and recognize the complexity and heterogeneity of the process (Goodwin, 2018;Levien, 2007;Li, 2014;Nair, 2019).
In addition to making an original contribution to debates about Polanyian social theory, this article also provides fresh insight into land markets and agrarian change. The existing scholarship on the Ecuadorian case provides considerable insight into land markets and struggles at the local level (e.g., Forster, 1989;Martínez, 1985;Thurner, 1989).
However, there has been no serious attempt to draw out broader patterns and consider how land markets evolve over time. Moreover, the dual role land markets performed for highland indigenous peasants in the late 20th century-simultaneously creating and reducing opportunities to obtain land-has not been properly explored or conceptualized. Although I use the activation-development concept to explore a specific social and historical context and make no claims about its wider validity, it might offer broader insights into land markets, especially in Latin America where colonial legacies, indigenous struggles and peripheral capitalism have fused to create distinct agrarian trajectories. The analysis I present in this article also contributes to debates about rural land and agrarian change elsewhere in the Global South (e.g., Chimhowu & Woodhouse, 2006;Li, 2014;Lavers, 2018).
The rest of this essay is structured as follows. I start by briefly explaining the links between land reform and land markets in Highland Ecuador, showing how land reform simultaneously stimulated and regulated land market activity. I then explore indigenous participation in land markets in the late 20th century, arguing that the activation of land markets opened spaces for highland indigenous peasants to obtain land, whereas the development of land markets closed them down. I conclude by connecting these historical processes to contemporary developments and drawing out broader implications for land struggles and agrarian change in the early 21st century.

| LAND MARKETS AND LAND REFORM IN HIGHLAND ECUADOR IN THE LATE TWENTIETH CENTURY
The fictional status of land as commodity obligates capitalist states to create, maintain and regulate land markets (Polanyi, 1944(Polanyi, /2001. Hence, they do not emerge spontaneously through exchanges between individual actors but require substantial state intervention to bring them into being and maintain them over the long run. This includes dissolving or weakening alternative forms of exchange and creating and maintaining a legal and bureaucratic framework to protect and promote private property rights and facilitate and regulate market exchange. Thus, capitalist states perform a dual role in land commodification, simultaneously promoting and regulating the process (Goodwin, 2017).
In the late 20th century, the Ecuadorian state used various mechanisms to perform this dual role, but land reform was the most decisive. Reform, which started in 1964 and ended in 1994, involved the abolition of traditional peasant-patron practices and the redistribution and regulation of agricultural and pastoral land. Its direct impact was limited: Only around 12% of the highland agricultural land surface was redistributed (Goodwin, 2017). However, its indirect effect was enormous, gradually eliminating traditional peasant-patron practices, dismantling the traditional hacienda complex and accelerating the capitalist modernization of agriculture.
One important feature of this agrarian transformation was the expansion of land markets. Land reform triggered a series of interconnected processes that enabled the market to perform a greater role in the distribution and valuation of land, while also generating contestation over the reach and operation of land markets. The rest of this section is dedicated to explaining these processes, focusing on the five dimensions outlined in Table 1. The primary aim is to provide the backdrop for the analysis of the activation and development of land markets in the next section.
However, this section also provides insight into the relationship between land reform and land markets, a topic that has received relatively little scholarly attention outside the market-based land reforms that were introduced across the Global South in 1990s and 2000s (Borras, 2003;Cottyn, 2018).
First, the central aim of Ecuadorian land reform was to dissolve traditional peasant-patron practices and accelerate the capitalist modernization of agriculture (Bretón, 1997;Chiriboga, 1988;Comité Interamericano de Desarrollo Agrícola [CIDA], 1965). Eliminating these practices ruptured the traditional hacienda complex and stimulated peasant demand for land through the market. The most common and important traditional practice in the highlands, huasipungo, involved peasant families exchanging their collective labour power for usufruct rights to land. A combination of social, political and economic factors combined to weaken huasipungo in the 1950s, and some landowning elites started to sell land to peasants to dissolve the practice at this stage, especially in the northern highlands, which were close to rapidly expanding domestic markets (Barsky, 1988;Becker, 2008;Guerrero, 1984). Land reform then generalized this process and helped erode huasipungo and other peasant-patron practices in the 1960s and 1970s.
Thus, from the late 20th century, the traditional route for highland peasants to secure land was steadily closed, making them increasingly reliant on land markets. The size and quality of the plots most former huasipungueros obtained set limits on their capacity to redistribute land to family members over the long run (Martínez, 2002;Muratorio, 1980). 6 The scarcity of communal land also meant most community authorities lacked the capacity to redistribute land to community members (Ministerio de Agricultura y Ganadería [MAG], 1977;Martínez, 2002).
T A B L E 1 Land reform and land markets in Highland Ecuador   Regulation of private property and land markets Regulated land market activity and created a site of struggle over the reach and operation of land markets Although many indigenous communities secured collective titles to land through land reform, the total stock of communal land remained limited. Moreover, the bulk of this land was páramo, high-altitude grassland, which is generally not suitable for cultivation, especially over the long run. 7 Meanwhile, the limited amount of land redistributed through land reform increased demand through the market, particularly in regions without state-owned haciendas. 8 Hence, when land reform accelerated the dissolution of the traditional hacienda complex, peasants faced difficulties obtaining land through family, community and state institutions and became increasingly dependent on land markets to maintain their rural livelihoods.
Second, although private property rights were widespread before land reform commenced, private land ownership spread as peasant-patron practices unravelled and land titles proliferated. 9 Usufruct rights were converted into private property rights, and formal land titles were issued through land reform. This included collective titles granted to communities and cooperatives; however, the bulk were issued to families. 10 Moreover, most collective titles issued to cooperatives were ultimately converted into private titles (Martínez, 1995(Martínez, , 1998Martínez & Martínez, 2019;Santana, 1995). Meanwhile, some communal land was also formally divided into individual plots, sometimes with the active involvement of state authorities (Forster, 1989;Martínez, 2002). Land titling took decades to complete and thousands of peasants lacked formal land titles when land reform ended in the 1990s (FAO, 1995(FAO, , 2002. 11 Yet, private land titles steadily increased in the late 20th century, which facilitated the exchange of land through markets. Third, land reform legislation conditioned private property rights on the economic use of land, which encouraged elites to divide haciendas, sell underutilized or unproductive land and prioritize the economic dimensions of land. This marked a radical break with the traditional agrarian economy. The traditional hacienda was a 'total social fact' that encapsulated virtually every aspect of social life, including the cultural, religious, political and economic (Mauss, 1925(Mauss, /2002. Thus, rural land performed multiple functions for elites, granting them social status, political power, religious authority and economic stability (Lyons, 2006;Muratorio, 1980;Sylva, 1986). Conditioning private property rights on economic use signalled the transition to an overtly capitalist agrarian economy that demanded greater capital investment, intense land and water use, and increased agricultural production for domestic and international markets (Bebbington, 2004). Land reform legislation issued in the early 1970s moved most decisively in this direction, with the law demanding landowners use at least 80% of land efficiently and reach average productivity levels or risk expropriation (RO 410, 1973). Supplementary legislation, bureaucratic ineffectiveness and elite capture prevented these measures from ever being properly enforced, and little land was actually expropriated and redistributed through the application of these clauses (Barsky, 1988;Chiriboga, 1988;Goodwin, 2017). Yet linking private property rights to economic use encouraged elites to prioritize the economic dimensions of land, which stimulated its exchange through markets.
Fourth, land reform supported indigenous and peasant organization and this increased the pressure on hacendados to sell land. Although some elites happily sold land to peasants and reaped healthy profits in the process, others were less enthusiastic and attempted to preserve their traditional status (Cervone, 2012;Guerrero, 1984;Sylva, 1986). Indigenous and peasant mobilization was required to force recalcitrant elites to sell. Indigenous and peasant communities, which multiplied under land reform, created a territorial and organizational base for this mobilization (Korovkin, 1997;Macas, Belote, & Belote, 2003;Zamosc, 1995). Community councils (cabildos) provided a mechanism to mobilize peasants within and between communities and coordinate mass occupations and protests.
Meanwhile, in the early 1970s, the establishment of the highland indigenous movement, Ecuarunari, and the 7 Census data indicate approximately 10% of the highland agricultural land surface was under communal control in the early 1970s, nearly 50% of which was classified as paramo and less than 15% deemed cultivable (INEC,1974). 8 State-owned land was estimated to amount to 132,853ha at the start of land reform (CIDA,1965). The state secured the bulk of this land through the expropriation of haciendas owned by the Catholic Church at the start of the 20th century. For insight into the redistribution of this land, see Goodwin (2017). 9 Agricultural census data give a sense of this increase: the proportion of private landholdings in the highlands increased from two-thirds to three-quarters between 1974 and 2000, whereas the absolute number of private landholdings nearly doubled (INEC,1974(INEC, ,2000. 10 In most cases, land titles were issued to male heads of households. On the gender dimensions of Latin American land reforms, see Deere (1987 radicalization of the national peasant movement, Federación Nacional de Organizaciones Campesinas (FENOC) provided peasants with a stronger platform to pressure the elite and the state (Huarcaya, 2018;Sánchez-Parga, 2010;Santana, 1995). The Federación Ecuatoriana de Indios, which was established in the 1940s, also remained active in local land struggles until the early 1980s (Becker, 2008;Cervone, 2012;Goodwin, 2017). In the late 1980s, the baton was passed to the Confederación de Nacionalidades Indigenas del Ecuador (CONAIE), which incorporated Ecuarunari into a national movement and breathed new life into indigenous land struggles as the 20th century drew to a close (Pallares 2002;Benavides, 2004;Cervone, 2012).
Fifth, although land reform stimulated land commodification, it also placed some restrictions on the operation of land markets. Regulation, which peaked in the early 1970s, included overseeing the division, transfer and sale of agricultural land, both inside and outside the reform sector. For example, officially, land reform beneficiaries required authorization from the Instituto Ecuatoriano de Reforma Agraria y Colonization (IERAC) before dividing and exchanging their land (RO 297, 1964;RO 410, 1973). From the early 1970s, technically, IERAC was also responsible for overseeing the division of rural landholdings to ensure it remained within 'the spirit of the law' (RO 410, 1973). However, these regulations were poorly enforced and easily circumvented, meaning actual land market regulation was significantly weaker than implied by the formal legal and bureaucratic framework. Indigenous and peasant movements repeatedly pushed for greater regulation through the enforcement of existing legislation and the introduction of new land laws (Goodwin, 2017). The state's repeated failure to respond to their demands supported the uneven accumulation of land through the market and deepened the inequalities that were created or embedded through the partial implementation of land reform.
Land market regulation weakened further in the 1990s after the introduction of Ley de Desarrollo Agrario. The law, which was actively promoted by international development agencies, boasted the 'sale and purchase and transfer of privately-owned agricultural land are free and require the authorization of no one' (RO 55,1997). The division of communal land was also facilitated, even if mass indigenous mobilization forced the Durán Ballén government to revise its original proposals and increase the proportion of community members required to vote to divide communal land (excluding páramo) to two-thirds (RO 55, 1997;RO 315, 2004;Bretón, 1997;Martínez, 1998;Goodwin, 2017).
The introduction of Ley de Desarrollo Agrario started a new phase of land commodification as domestic and global capital poured into the agricultural sector and the economic valorization of land intensified (FAO, 1995;Guerrero, 2004;Martínez, 1998). Yet, viewed in historical perspective, the law was one step in a longer term process of land commodification, which started to accelerate after the introduction of land reform in the 1960s.
To sum up, land reform both stimulated and constrained land markets but moved far more decisively in the direction of the former than the latter, meaning the activation and development of land markets took place within a weak regulatory framework that favoured landowning elites, medium-sized farmers and agroindustrialists. The next section explores this process, explaining the spaces that opened and closed for highland indigenous peasants to purchase land as the capitalist modernization of agriculture advanced.

| INDIGENOUS PEOPLES AND LAND MARKETS IN THE LATE TWENTIETH CENTURY: ACTIVATION AND DEVELOPMENT
Land reform enabled the market to perform a greater role in the use, valuation and distribution of land in the late 20th century; however, it also converted the land market into a site of social and political struggle. Indigenous efforts to secure land through the market were emblematic of this broader process. In this section, I explore these struggles through the lens of the activation-development concept outlined in the introduction to this article. Drawing on several cases, I argue that the activation of land markets opened spaces for indigenous peasants to acquire land, whereas the development of land markets closed them down. I also show that social and class differentiation increased among the highland indigenous population through unequal participation in the land market.
The section starts by briefly outlining highland indigenous land relations in the late 20th century, explaining the role family and community institutions performed in regulating the exchange of land and the factors that contributed to the fictional status of land as commodity for indigenous peasants. The analysis then turns to indigenous efforts to purchase land through the market, starting in the 1960s and 1970s before moving on to the 1980s and 1990s.

| Highland indigenous land relations in the late twentieth century
Highland indigenous land relations took specific social, historical and political forms in the late 20th century. Land tenure on rural indigenous communities was largely based on private ownership, with indigenous families securing plots of land through various mechanisms, including land reform and land markets. Family plots were generally divided and subdivided between family members through inheritance and marriage (Belote, 1984;Martínez, 2002;Sánchez-Parga, 1984). The timing and mode of the transfer of land depended on various factors including, local customs and practices, the amount and quality of land available, the size and age of the family, and the productive strategies of the family unit. Land was generally redistributed among sons and daughters sometimes before but usually after the death of the head of the household. 12 Before land was definitively transferred through inheritance, adult children generally lacked formal land titles but usually enjoyed access to family land, which strengthened relations of dependence and reciprocity across generations and increased the productive capacity of the family unit. Yet the limited amount of land available to most indigenous families made it difficult for them to absorb family labour over the long run, which encouraged young indigenous peasants to turn to the market to obtain land. Marriage provided further opportunities for them to access land. Where sufficient land was available, heads of households usually ceded small plots to young married couples, although newly formed families generally lacked autonomous control of the land until it was formally transferred through inheritance. The allocation of land through marriage sometimes provided opportunities for extended indigenous families to gain access to land in different ecological zones, minimize the risk of crop failure and diversify agricultural production. However, this traditional Andean practice proved difficult to sustain as family plots became increasingly fragmented. The distribution of land through marriage also strengthened ties between families within and between communities, enmeshing indigenous peasants within a broader web of family and community relations.
In some cases, members of highland indigenous families also enjoyed usufruct rights to communal land, especially páramo, which was mostly used for pasture and sometimes for cultivation (MAG, 1977;Martínez, 1998).
Hence, private and communal land tenures often coexisted on highland indigenous communities (Martínez, 2002;Pacari, 1993). However, as noted above, communal land was scarce, and many indigenous communities lacked collective titles to agricultural or pastoral land. 13 Yet, even in these cases, community institutions still influenced the distribution of land, with the buying and selling of private land generally, if not universally, regulated through communal relations and authorities (FAO, 1995). The sale and purchase of land in indigenous communities took multiple forms and included various actors. In most cases, it involved transactions between members of the same community, with community members deciding to sell land for diverse reasons, including permanent migration and economic distress.
Land was also sometimes sold to members of the same family following the division of family plots through inheritance. In other cases, indigenous peasants exchanged land with members of neighbouring indigenous communities or occasionally sold land to mestizo peasants or middlemen (Belote, 1984).
Ley de Comunas, which was first introduced in 1937, supported the communal regulation of land (RO 558, 1937). 14 To register as a comuna, communities were required to establish a community council (cabildo) and declare collective goods (bienes colectivos or bienes en comun), including, where applicable, communal agricultural and 12 For a vivid visual representation of the fragmentation of land in a highland indigenous community in the 20th century, see Iturralde (1980, pp. 125-129).
See also Sánchez-Parga (1984, pp. 168-170). 13 For example, Martínez (2002, p. 41) reports that in the late 1970s, 250 of 1,318 legally registered highland peasant communities owned communal land, and the bulk of this land was divided into family usufructuary plots. 14 For further insight into indigenous engagement with the Ley de Comunas, see Santana (1995), Becker (1999), Martínez (1992Martínez ( ,2002, Lucero (2003) pastoral land. 15 The law stated that bienes colectivos belonged to all members of the community and charged the cabildo with the task of governing them. Hence, the cabildo's main task was to oversee communal property. However, as Santana (1995) notes, the legal framework provided plenty of scope for intervention in other areas. For example, the cabildo was expected to defend the territorial integrity of the community, which created space for the regulation of private as well as communal land. Moreover, indigenous communities tended to blend the formal requirements of the law with local customs and practices to create distinct, vernacular forms of communal organization (Colloredo-Mansfield, 2009;Korovkin, 1997;Rayner, 2017). 16 Thus, the communal authorities and relations that regulated land varied between indigenous communities. The degree of communal influence over the exchange of private land depended on multiple factors, including the internal statutes and rules of the comuna and the power and legitimacy of the cabildo. In some cases, individuals and families enjoyed considerable autonomy, whereas in others community authorities exerted greater influence. In the 1990s, the Ley de Desarrollo Agrario challenged the cabildo's authority to regulate the exchange of private land and created greater space for the division of communal land (RO 55, 1997;RO 315, 2004). However, the Ley de Comunas provided indigenous communities with a tool to counteract the liberalization of the land market and prevent or reduce the division of communal land (Martínez, 1998;Rayner, 2017). Hence, indigenous communities were able to use the law to frustrate state efforts to erode the communal regulation of land and expand the free exchange of land on the market.
Historical and political factors added to the uniqueness of indigenous land relations and supported the family and community regulation of land in the late 20th century. Colonialism involved the steady dispossession of indigenous land, and this process weighed heavily on indigenous land politics (Benavides, 2004;Huarcaya, 2018;Meisch, 1992;Pacari, 1993). The lived experiences of indigenous peasants on the traditional haciendas that emerged out of colonialism were fundamental to indigenous collective memories. For example, describing life on a traditional hacienda in Guamote, Chimborazo, one indigenous peasant told me 'my mother and father could not obtain land so they worked the land of the masters (amos), they were slaves of the masters (amos), serving them, cleaning their backsides, cleaning their land, cleaning their buildings'. 17 Collective memories of exploitation and oppression fused with shared experiences of poverty, inequality and racism to provide a powerful basis for collective organization and mobilization. Building on these memories and experiences, local and national indigenous leaders increasingly stressed cultural distinctiveness and ethnic identities to strengthen collective organization and fuel land mobilizations (Benavides, 2004;Cervone, 2012;Sánchez-Parga, 2010). The centrality of land to highland indigenous culture supported these efforts, and land and cultural demands became increasingly enmeshed as the century drew to a close. In turn, this underpinned attempts to build local indigenous autonomy and transform relations with the state.
Hence, land was central to a wider indigenous political project, which crystallized at the end of the 20th century in demands to transform Ecuador into a plurinational state (Becker, 2012;Benavides, 2004;Rayner, 2017). 15 See Lentz (1997, pp. 254-282) for an ethnohistorical analysis of the formation and evolution of a highland indigenous comuna in the 1970s and 1980s. 16 Other indigenous communities shunned the law and remained outside the legal framework or took decades to register with the authorities (Becker,1999). See also Santana (1995), Colloredo-Mansfield (2009) 3.2 | Highland indigenous land market struggles in the late twentieth century Two cases examined by William Waters (2007) in the northern highland province of Cotopaxi highlight the diverse opportunities that emerged for indigenous peasants to purchase land under land reform. The first case revolved around the division and dissolution of a traditional hacienda once owned by the former conservative president of Ecuador, Gabriel Garcia Moreno (Waters, 2007, pp. 131-134). Descendants of the former president sold most of their remaining land when land reform surged up the political agenda in the early 1960s. Reflecting broader changes in the highland agrarian economy, Waters reports that the bulk of this land was sold to a new class of landowner rather than the traditional landowning elite. The rest was granted or sold to indigenous peasant families who had worked on the hacienda as huasipungueros. Land was sold at the family level. However, the indigenous families organized into a community, following the general pattern in the highlands. Along with a neighbouring community, the families also secured collective ownership of 21,000 ha of páramo through negotiation with the land reform agency, IERAC (Waters, 2007, p. 133). Through the subdivision of the hacienda in the 1960s and 1970s, indigenous peasants obtained small plots of good quality land for cultivation as well as collective ownership of a huge tract of páramo for pasture. Yet, as Waters notes, 'a second generation of large-and medium-sized landowners emerged from the same process and continued to occupy substantial amounts of highly productive land that remained out of the hands of the indigenous community' (Waters, 2007, p. 134).
Indigenous families involved in the division of the second hacienda analysed by Waters were less fortunate (Waters, 2007, pp. 134-137). This hacienda was passed down through a wealthy highland family, before being sold in the late 1940s to two urban business partners who took equal shares. In the mid-1960s, one of the partners liquidated huasipungo through the transfer of land to former huasipungueros, who created a small community. Indicating some differentiation among the indigenous families, Waters reports that the 'parcels ranged in size from 0.3 to 1.9 hectares, averaging only 1.42 hectares' (Waters, 2007, p. 135). During the next decade, the families augmented the small pieces of land they obtained through the dissolution of huasipungo with the purchase of individual plots of cultivable land and the collective acquisition of 1,529 ha of páramo. Thus, the land market provided them with opportunities to secure land and combine private and communal land tenures. However, the cultivable land they purchased was elevated and unproductive, receiving 'virtually no irrigation water … and dominated by a heavy, unproductive clay soil' (Waters, 2007, p. 135). The best quality land was retained by the original hacendados and a veterinarian from the nearby commercial city, Ambato, who purchased 349 ha in the 1970s. The outcome, as Waters explains, was that the haciendas monopolized 'not only the better lower lands, but also the most productive higher elevation lands' (Waters, 2007, p. 135).
Despite the different outcomes in the two cases, a similar pattern emerged. On the one hand, the breakdown of the traditional hacienda and the activation of the land market created opportunities for indigenous peasants to secure small plots of elevated cultivable land and large tracts of high-altitude pastoral land. On the other, the development of the land market restricted indigenous access to the most productive land located at lower elevations, with this land being sold to a new generation of landowning elite or retained by the original hacendados. Hence, the activation and development of land markets exhibited a distinct geographic pattern: indigenous peasants were generally able to purchase more marginal land at higher elevations but were locked out the market for the best quality land located along the floor of the Andean valley.
The case studied by Carola Lentz (1997) in the central highland province of Chimborazo paints a similar, if more heterogenous, picture. The diversity of outcomes was linked to the land purchases that some indigenous peasant families realized before land reform, with the income indigenous workers earned from temporary work on coastal sugar plantations supporting land acquisitions, especially from the 1940s (Lentz, 1997, pp. 51-53). Lentz notes that these early land purchases, which were facilitated by low market prices, were not the atomized, impersonal transactions imagined by neoclassical economic theory. Rather, 'they depended on a complex network of social relations: the offer, the price, and the methods of payment depended to a large extent on the specific relations between the community buyers and the landowner' (Lentz, 1997, pp. 51-52). Personal relations therefore contributed to the heterogeneity of early land market outcomes, with some indigenous peasants exploiting their relations with the hacendado and purchasing land on more favourable terms.
Lentz reports that the build-up to land reform in the early 1960s accelerated the dissolution of the hacienda and increased the amount of land offered for sale (Lentz, 1997, pp. 54-59). The landowner's decision to liquidate huasipungos before land reform facilitated his efforts to sell rather than grant land to huasipungueros: only two huasipungueros received land in exchange for their labour on the traditional hacienda; the rest were evicted from their landholdings and sold plots of more marginal land. Describing this process, one former indigenous worker recalls: 'In the end, people started to buy but they did not sell those huasipungos in the pamba. Where we are now, on the slopes, here and down there they started to sell' (Lentz, 1997, p. 54).
Land reform then accelerated the division of the hacienda. Lentz estimates a quarter of the land offered for sale in the late 1960s was purchased by members of the indigenous community, with the rest purchased by members of neighbouring communities or mestizos from a nearby town (Lentz, 1997, p. 56). However, the operation of the land market started to work against the indigenous peasants in the late 1970s. Their refusal to work on the hacienda and their decision to form a legally registered comuna contributed to the hacendado's decision to sell the remaining sections of the hacienda 'piece by piece' (Lentz, 1997, p. 57). Yet the landowner took advantage of the tightening of the land market and offered land for sale at significantly higher prices. Lentz estimates that the prices the hacendado demanded for the remaining sections of the hacienda were 50 times greater than the prices in the 1960s, whereas the wages the indigenous workers received on coastal sugar plantations were only 10 times higher (Lentz, 1997, p. 57). With few savings and limited incomes, indigenous peasants were effectively priced-out of the market. Hence, the development of the land market progressively closed-down spaces for indigenous peasants to obtain land.
The process described by Lentz occurred elsewhere in the highlands as landowners exploited the market to maximize economic returns, frequently pitting one group of buyers against another to escalate prices (Huarcaya, 2018;Martínez, 1984Martínez, , 1985Muratorio, 1980;Thurner, 1989Thurner, , 2000. Luciano Martínez (1985), for example, shows the difficulties increasing land prices caused indigenous peasants in a small community in the central highland province of Chimborazo. He argues that, although many indigenous families purchased small plots of land, the land market ultimately 'benefitted the landowners who bowed to peasant demands but obtained significant earnings from the sale of their properties' (Martínez, 1985, pp. 152-153). Hence, the exchange of land through the market embedded economic inequalities between indigenous peoples and landowning elites. However, the obstacles in front of another indigenous community he analyses in the northern highland province of Imbabura were far greater. There, land prices largely, if not totally, blocked indigenous efforts to purchase land as medium-and large-sized haciendas monopolized the best quality land (Martínez, 1985, pp. 154-156).
Market prices were also a considerable obstacle to the indigenous families I interviewed in a small community in Chimborazo. 18 The community members explained that it was only during the build-up to the introduction of the second major land reform law in the early 1970s that local hacendados started to sell land to indigenous peasants on a significant scale. Indigenous mobilization created additional pressure as strikes, protests and occupations occurred across the province, creating 'significant concern' among local elites. The introduction of the law in 1973 and the escalation of indigenous resistance prompted more elites to start selling land to indigenous peasants, most of whom purchased small plots.
During this process, the community members were initially offered remote land located high on the slopes of the Andean valley. However, this land was not suitable for cultivation. Instead, they purchased small plots of cultivable land lower down the valley but still a strenuous two-hour walk along a rocky path from the town centre and local market. Each family pooled their resources and paid a small deposit and a series of annual instalments to the hacendado. One community elder said that he later purchased another small extension of land from the daughter of the hacendado, using resources secured through his extended family and wage labour: 'we fixed the price, we paid it, and we live in peace' . Yet market prices set definite limits on the quality and quantity of the land the indigenous families could purchase. The community members said that indigenous peoples did not have the money to purchase the most productive land incorporated into the hacienda so the hacendados divided it between family members with 'great ease' . By way of example, one community member explained: 'Dr Francisco Martínez, he took the section opposite called San Rosa, and the sister Elena Martínez the section behind, another called Alfonso Martínez and the wives of some hacendados, they took the section up to the bridge'. 19 Hence, through the activation of the land market indigenous families secured small plots of relatively marginal land. The best quality land, which remained under elite control in the 1960s and 1970s, was available to purchase on the market but at a price that far exceeded the economic capacity of indigenous families.
In this case, it was the broad build-up of indigenous and peasant mobilization that prompted landowners to start to sell land. The cases studied by Mark Thurner (2000) in Chimborazo illustrate the more direct impact of indigenous mobilization on land markets. 20 Echoing Lentz, he describes how indigenous workers used the income they generated from toiling on coastal sugar plantations to purchase land and reduce their dependence on the hacendado prior to land reform (Thurner, 2000, pp. 364-365). Land reform then accelerated this process as a new generation of indigenous leaders established a comuna to 'defend and expand their control of land at the expense of the hacienda' (Thurner, 2000, p. 366). However, although collective organization was central to land acquisitions, indigenous land purchases were highly differentiated. Some secured more land than others and employed other indigenous peasants to work their land. Hence, participation in the land market contributed to class differentiation among indigenous peoples.
Events on a neighbouring hacienda followed a different trajectory. The estate was not divided until 1965 when it was split in two, with each half being sold separately in 1972: 'One half went to the old mayoral and mayordomo while the other half was sold to a mestizo compadre from a nearby village' (Thurner, 2000, p. 367). Following the general trend, these land sales reflected the ethnic division of labour on the traditional hacienda, that is, the best quality land was often purchased by mestizos who held more powerful and well remunerated positions (e.g., Weismantel, 1989). However, indigenous mobilization and resistance performed a crucial role in derailing the commodifying plans of one of the new landowners. His later attempts to sell pastoral land attached to the hacienda were blocked by members of various local communities who occupied the land and ultimately secured communal ownership of around half of it through the intervention of the land reform agency, IERAC (Thurner, 2000, pp. 348 and 368). The development of the land market therefore locked out indigenous peasants from the most productive sections of the hacienda. However, a combination of collective mobilization and state intervention blocked the full incorporation of pastoral land into the land market.
These cases show that the division of traditional haciendas and the capitalist modernization of agriculture in the 1960s and 1970s saw the exit of some landowners and the entrance of others (Barsky, 1988;Chiriboga, 1988;Sylva, 1986;Waters, 2007). The new owners of the medium-and large-sized landholdings that emerged through this process included the family members of traditional landowning elites, former hacienda administrators and managers, and companies and investors attracted to the profits available from capitalist agriculture and land speculation.
Stripped of many of the non-economic functions land performed for elites in the traditional agrarian economy, the economic dimensions of land became more central for this group of landowners, with the economic valorization of land being realized through the market. Opportunities for indigenous peasants to acquire productive land were limited and most were restricted to purchasing relatively small plots of marginal land located at higher elevations.
Hence, although traditional haciendas disappeared and medium-sized farms proliferated, the land market embedded underlying ethnic land inequalities, with the bulk of indigenous peoples remaining at the lower end of the land distribution in the 1960s and 1970s (Goodwin, 2014 notes, for example, that land sales to indigenous peasants in Chimborazo, which remained highly heterogenous in the 1980s, were 'the indirect result of the social and political threats of invasion and expropriation made concrete in labor indiscipline and other forms of local resistance' (Thurner, 1989, p. 29). Emma Cervone provides further evidence of this, with indigenous families in Tixan using occupations to persuade elites to sell them land (Cervone, 2012, pp. 139-143). Indicating the power of this collective action and the political subjectivities it fostered, she notes that 'impressive numbers of indigenous peasants met on the seized properties to support and encourage their comrades at different stages of the struggle' and 'land seizures became a rite of passage underscoring the birth of new political subjects' (Cervone, 2012, pp. 142 and 144). Strikes, protests and occupations continued across the highlands in the 1980s and 1990s, with indigenous peasants maintaining the pressure on landowners to sell as the possibilities of obtaining land through land reform became ever more remote (Goodwin, 2017;Guerrero, 2004;Huarcaya, 2018).
The emergence and strengthening of the national indigenous movement, CONAIE, played a pivotal role in this resistance, with the movement spearheading national mobilizations and supporting local land disputes in the 1980s and 1990s (Rosero, 1990). The 1990 indigenous levantamiento-uprising-was particularly effective in unlocking land markets and increasing indigenous control of land (FAO, 1995;Navarro, Vallejo, & Villaverde, 1996). The levantamiento, which started with the occupation of the iconic Santo Domingo church in Quito, swept across the highlands in early June 1990, causing considerable consternation among landowning elites, especially in Chimborazo where land disputes were most intense and widespread (El Comercio, June 1, 1990: A-6; El Espectador, June 6, 1990: 1-8; Rosero, 1990;Meisch, 1992;Zamosc, 1994;Korovkin, 1997). 21 The 1994 levantamiento, which was launched to challenge the introduction of Ley de Desarrollo Agrario, generated additional pressure and ensured land remained high on the social and political agenda (El Comercio, June 3, 1994: 1;Bretón, 1997;Sawyer, 2004;Goodwin, 2017). Some landowning elites stood firm against escalating indigenous pressure, but many decided to sell, including productive land that was previously out of the reach of most indigenous families and communities.
Yet, while collective organization enabled indigenous peasants to secure land through the activation of the land market, the sharp economic downturn of the 1980s and 1990s complicated their efforts. On one hand, surging unemployment and declining real wages undermined attempts to channel income from wage labour into land purchases. On the other, rising agricultural input prices and falling demand for agricultural produce reduced incomes from agricultural production. Hence, real incomes declined inside and outside the agricultural sector, limiting possibilities to purchase land (Cervone, 2012;Jokisch, 2002).
The impact of the economic slump was compounded by escalating land prices as the land market continued to tighten. FENOC, the national peasant movement, captures the dilemma facing peasants in the early 1980s: Taking advantage of the scarcity of productive land, the understandable desire of the peasant to acquire land, and the demographic pressure in minifundista zones, landowners have attempted to sell their land at the price of gold … The growing imbalance between the price of land and the income of minifundistas progressively diminishes the possibilities of transformation. (FENOC, 1984, p. 9) Various factors contributed to escalating land prices during this period. First, the basic imbalance between supply and demand placed general upward pressure on prices, especially as the rural population continued to grow despite rapid urban migration (Forster, 1989;Korovkin, 1997;Thurner, 1989). Notably, pressure was most acute in zones with high indigenous populations, which generally experienced less permanent migration (Kramkowski, 1989;Zamosc, 1995). Second, revisions to land reform legislation in the late 1970s and early 1980s strengthened the protection of private property rights, which reduced the threat of expropriation and increased the economic value of 21 Interviews: Indigenous leader, December 14, 2018, Guamote; Regional Coordinator, FEPP, Riobamba, August 5, 2010.
This last factor was particularly influential in shaping the dynamics of land markets in the highlands in 1980s and 1990s. Historically, agricultural export production in Ecuador was concentrated in the coastal region, whereas production for the domestic market was located in the highlands. However, the frontiers of the agroexport sector expanded in the 1980s and 1990s as the unravelling of land reform and the liberalization of trade and capital flows encouraged greater investment in agroexport (FAO, 1995;Korovkin, 2004;Rubio, 2008). This was particularly evident in the northern highland provinces of Cotopaxi, Imbabura and Pichincha, where highly productive land, extensive irrigation networks, favourable climatic conditions and good transport links attracted agroexporters in their droves.
The rapid expansion of the cut flower industry was the starkest illustration of this wider trend. Flower plantations started to spread in the late 1980s, before expanding rapidly in the 1990s and 2000s (El Comercio July 15, 1988: A-6;FAO, 1995). Elevated land prices and considerable capital and technological requirements precluded the involvement of indigenous families as producers; the only space open to them to participate was as wage labourers on flower plantations (Korovkin, 2004;Martínez, 2013;Rubio, 2008). 22 Flower plantations were emblematic of the capital-intensive medium-sized farms that emerged in the highlands as the traditional hacienda unravelled, with most ranging between 20 and 50 ha (Instituto Nacional de Estadísticas y Censos [INEC], 2000). Crucially, however, land incorporated into the flower sector was among the most productive and best irrigated in the highlands, revealing a new form of land inequality in which the quality rather than the quantity of land was the key criteria. The basic pattern of land ownership that emerged in regions where flower plantations expanded was the monopolization of the most productive land located at lower elevations by national and international flower producers and the restriction of indigenous and mestizo peasants to more marginal and less productive land at higher elevations (FAO, 1995, p. 54). Hence, the sector reflected the broader social and spatial pattern of land ownership that emerged across the highlands in the 1960s and 1970s. Indigenous peasants still purchased land through the activation of the land market in the 1980s and 1990s. However, the development of the land market and the growth of the agroexport sector generally prevented them from obtaining the most productive land.
The Fondo Ecuatoriano Populorum Progressio (FEPP) land purchase programme provides further insight into this process (Bretón, 2008;Cervone, 2012;Navarro et al., 1996). The initiative, which supplied interest-bearing loans to peasants to purchase land on the market in the early 1990s, responded to two main factors: the inability of peasants to secure land via the market or land reform and the upsurge in rural land conflicts, social unrest and popular protests. The 1990 levantamiento forced religious authorities and state officials to act. Shortly after the uprising, the Conferencia Episcopal Ecuatoriana reached agreement with the Borja government to purchase discounted public overseas debt, convert the debt into national currency and invest it into social programmes (Navarro et al., 1996, pp. 280-282). One of the most important schemes was the Programa Nacional de Servicios Integrados a las Comunidades Indígenas, which received around a third of the total funds, 60% of which was channelled into the purchase and legalization of land (Navarro et al., 1996, p. 62). The FEPP land purchase programme was part of this initiative. 23 Landowning elites lobbied hard to ensure the programme protected or extended their economic interests, including visiting FEPP offices and meeting state officials. This individual and collective pressure was reflected in the inclusion of regulations that demanded 'free and voluntary' sales of land and purchases at 'market prices without subsidies' (Navarro et al., 1996. p. 73). Hence, the programme resembled the 'willing buyer, willing seller' marketbased land reform model promoted by the World Bank in the 1990s (Borras, 2003). But it did not follow the economistic logic of this model. Indigenous and peasant pressure was required to force elites to sell and increase the supply of land on the market. Moreover, the programme responded to decades of indigenous resistance and was therefore inherently historical and political.
A full analysis of the FEPP programme is outside the scope of this essay; however, inspection of some of its main characteristics provides further insight into the activation and development of land markets (Table 2). Three important consequences of using market prices as barometers for land value followed for indigenous peasants.
First, purchases took place in the context of rapidly increasing land prices as landowners exploited the imbalance between supply and demand to maximize returns (Navarro et al., 1996, p. 179). The programme contributed to this general trend by converting latent indigenous demand into effective demand (Rosero, 1990, pp. 61-62;Navarro et al., 1996, pp. 180-181). Elevated prices dissuaded some indigenous peoples from participating in the programme and complicated their efforts to secure land through the market without FEPP support. The scheme therefore contributed to differentiation among the highland indigenous population, with only a relatively small proportion willing and able to participate. Furthermore, although loans were granted at attractive interest rates and generally repaid quickly, elevated land prices saddled buyers with high levels of debt. (Navarro et al., 1996, p. 169). This sometimes reduced the cohesion of indigenous communities as community members migrated to secure income to repay the loans (Bretón, 2008). 24 Second, market prices limited the amount of land incorporated into the programme, especially in regions where capital-intensive agroexport took root. For instance, Navarro et al. note: 'In the province of Imbabura, the land market is restricted, and the prices are very high, especially in areas where there are enterprises that produce flowers for export markets ' (Navarro et al., 1996, p. 164). Land purchase data provide further evidence. The average price of a hectare of land purchased through the FEPP branch responsible for Imbabura was over four times higher than the average price of land purchased through the Chimborazo office (Navarro et al., 1996, p. 178). The former purchased 3.3% of the total amount of land incorporated into the FEPP programme but absorbed 11.5% of total funds, whereas the latter purchased 59.4% of land but absorbed 54% of funds (Navarro et al., 1996, p. 161). Hence, the development of the land market set definite limits on indigenous access to land in regions where agroexporters operated.
Third, market prices also influenced the quality of land. The combination of indigenous activism and interestbearing loans enabled indigenous peasants to purchase good quality land in some cases. For example, the FEPP representative I interviewed explained that the sections of the haciendas landowning elites decided to sell in the canton of Guamote were located along the floor of the valley where the most productive land is located. 25 The FEPP initiative enabled some indigenous peasants to obtain land that was previously out of their reach. One of the participants in the programme who I interviewed offered a similar account. 26  More broadly, however, the bulk of the most productive land remained out of the programme's reach. 29 Nearly two-thirds of land purchased through the scheme was located in Chimborazo, where land is typically less productive and more remote than in Cotopaxi, Imbabura and Pichincha (Table 2). Mirroring land reform, the land channelled through the initiative also included a significant amount of páramo (Goodwin, 2017). For example, most of the 10,950 ha of land acquired by one indigenous organization in Chimborazo comprised páramo (Navarro et al., 1996, pp. 150-152). This single transaction accounted for over one-third of the total amount of land purchased through the scheme in the highland region between 1990 and 1995. 30 Insistence on the collective purchase of land was another important feature of the FEPP programme. Collective organization was a precondition for participating, and credit was only extended to communities, associations and cooperatives. What's more, participants were required to work land collectively until the loans were repaid. New organizations were often formed to access loans and manage the land. With the price and amount of land restricting wider involvement, these organizations were established within existing communities as not all community members were willing or able to participate (Cervone, 2012;Navarro et al., 1996). In addition to contributing to social differentiation within indigenous communities, this added to the diversification of local-level indigenous organization, which tended to undermine traditional community authorities and complicate indigenous organization and mobilization (Bretón, 2015a(Bretón, , 2015bColloredo-Mansfield, 2009;Martínez, 1995;Sánchez-Parga, 1989).
Though well-intentioned, the requirement to purchase and work land collectively reflected a romantic vision of indigenous peasants that departed from reality in most highland indigenous communities in the late 20th century.
With the exception of páramo, as explained in the previous section, land was generally owned at the family rather than the community level. Indigenous families were bound together by a web of communal practices and relations; however, agricultural production was generally carried out within the family and little agricultural land was farmed at the community level, especially by the 1990s. Thus, the model of land tenure and agricultural production imposed by the programme was alien to most indigenous peasants. This helps explain why virtually all the organizations divided their land soon after repaying their loans (Navarro et al., 1996, pp. 209-211;Cervone, 2012, p. 76). Yet this land tended to remain within the community or family. The FEPP report that decisions to divide the land were not based on the desire to offer it for sale on the open market (Navarro et al., 1996, pp. 209-211). Inheritance was given as one of the principal reasons, indicating the continued importance of the family as a mechanism of land distribution.  29 On a national level, FEPP classified 10% of the land purchased through the scheme as 'highly productive', 61% as 'medium quality' and 29% as 'poor quality' (Navarro et al., 1996, p. 193). 30 Note that this transaction significantly skews upward the average land sizes reported in Table 2. 31 Interview: Regional Coordinator, FEPP, Riobamba, August 5, 2010.
undertake agriculture according to their own forms of organization, reflecting the impulse for autonomy among highland indigenous communities. Both the participants I interviewed reported a similar pattern and expressed similar sentiments, with virtually all the land remaining within their communities after it was divided. 32 Thus, land purchased through the FEPP programme tended to remain under indigenous control and regulated through family and community institutions.
To sum up, although some indigenous peasants obtained productive land through the FEPP programme, most of the best quality land remained out of their reach, especially where capital-intensive agroexporters operated. The initiative reflected the broader pattern that emerged in the highlands in the late 20th century, with the activation of land markets creating space for indigenous peasants to secure land and the development of land markets closing them down. The programme also increased social and class differentiation within indigenous communities, which was another key feature of this process. The outcomes of these historical processes have important implications for indigenous land struggles in the early 21st century. I turn to this point below. There and elsewhere, sustained collective organization and mobilization were required to unlock land markets and persuade elites to sell. The private land that indigenous peasants purchased was then typically, if not exclusively, circulated through family and community institutions rather than exchanged freely on the market. On the other hand, the development of the land market closed down spaces for them to obtain land as market prices played a progressively larger role in regulating the distribution and value of rural land. This process was evident across the highlands, but it was particularly acute in the northern provinces of Imbabura, Pichincha and Cotopaxi, where capital-intensive agroexport expanded in the 1980s and 1990s. Hence, the development of the land market was closely connected to the world economy.
Taking a longer-term view of the interaction between the activation and development of the land market shows the former playing a contradictory role in the latter. On one side, activation increased the amount of land under indigenous control and underpinned indigenous land mobilizations in the late 20th century. This challenged the development of land markets by increasing the amount of land circulated through family and community institutions and supporting a countermovement that attempted to decommodify land (Goodwin, 2017 distribution of land in the 2008 constitution, a coalition of indigenous and peasant movements-Red Agraria Nacional-developed a legislative proposal that sought to limit the reach and influence of the rural land market (Goodwin, 2017;Red Agraria Nacional, 2012). Yet the coalition excluded CONAIE, the chief protaganist in national land mobilizations in the 1980s and 1990s and the biggest indigenous movement in Ecuador, and little pressure was placed on the government to introduce a radical new land law. 34 Multiple factors contributed to CONAIE's absence from the coalition, including Correa's efforts to weaken and divide the indigenous movement, but outcomes of historical land struggles were pivotal, especially increased social and class differentiation within the highland indigenous population (Goodwin, 2017;Henderson, 2017Henderson, , 2018. Other factors have overlapped with these developments to reconfigure indigenous land relations and struggles in the early 21st century. The unequal distribution of land and water, the crisis of small-scale agriculture, and the allure of urban life have encouraged growing numbers of indigenous peoples to abandon agriculture and seek employment in towns and cities (Berry, Kay, Martínez, & North, 2014;Colloredo-Mansfield, 2009;Goodwin, 2019;Horn, 2018;Rayner, 2017). Highland indigenous peoples who live in urban areas generally retain strong links with rural families and communities; however, their connection to rural land and engagement in rural communal activities have weakened. In some cases, the direction of travel has been reversed as expanding towns and cities have come to rural indigenous communities, which has increased the economic value of land and created stronger incentives to offer it for sale on the market. 35 Tourism has also triggered rapid land price rises, stoking tensions within and between indigenous communities in some zones (Gascón, 2016). In other areas, tourism has created new economic opportunities for some rural indigenous families and communities, leading to a distancing from agrarian struggles and indigenous movements (Lyall, Colloredo-Mansfeld, & Quick, 2020). Participation in the agroindustrial sector has also influenced indigenous land relations and struggles. Offering the prospect of a relatively high and stable income, labouring on flower plantations has become commonplace for indigenous men and women in the northern highlands, even if the work is associated with serious health problems due to extreme working conditions and extensive use of chemicals (Korovkin, 2004;Martínez, 2013;Martínez & Martínez, 2019;Rubio, 2008). Elevated land prices in zones where flower plantations operate make it extremely hard for indigenous labourers to invest in productive land, whereas the crisis of small-scale agriculture offers few incentives to dedicate time and resources to cultivation. Indigenous engagement in the flower sector has also complicated the efforts of indigenous communities and movements to mobilize around land and added to social and class differentiation within the highland indigenous population (Soper, 2019).
The hybrid land regime that emerged in the late 20th century through the activation and development of the land market has been reconfigured through these changes. Highland indigenous peoples continue to circulate land through family and community institutions, including in urban areas in some cases (Rayner, 2017;Rayner & Merida, 2019). 36 However, distribution through the family has become even more difficult due to the lack of land available to most indigenous families and the extensive subdivision of land through marriage and inheritance (Martínez, 2002;Radcliffe, 2015). Meanwhile, community authorities and relations have come under strain due to increased social and class differentiation, temporary and permanent migration, and the proliferation of local organizations. The land market continues to exist alongside forms of exchange based around the family and community. Yet its role in regulating the distribution and value of land has become progressively larger as capitalist agriculture and relations have spread and private property rights have become more prevalent and entrenched.
Thus, the land market has become increasingly dominant in rural Highland Ecuador over the last two decades.
Yet land has not been stripped of its fictional status as commodity. It remains enmeshed in social, cultural and 34 In the event, the Correa administration introduced a new land law-Ley Orgánica de Tierras Rurales y Territorios Ancestrales-which was far removed from the Red Agraria Nacional proposal (RO 7112016). Although this legislation, which supersedes and annuls Ley de Desarrollo Agrario, provides protection for communal land, it places few meaningful restrictions on land markets. 35 Interview: Indigenous leader, Guamote, December 14, 2018. 36 The role family and community institutions perform in regulating the exchange of rural land was highlighted during my discussions with highland indigenous communities and leaders in the 2010s, especially in the province of Chimborazo.
political institutions, has multiple meanings and values and is instilled with basic properties that differentiate it from genuine commodities. However, the evolution of the land market and indigenous struggles show political tensions connected to the fictional qualities of land change over time. Recent literature on land relations and agrarian change elsewhere in the Global South points in a similar direction (e.g., Ferguson, 2013;Li, 2014;Naybor, 2015;Sud, 2019).
I suggest a simple modification to the fictitious commodities concept which seeks to improve its analytical power in light of these insights. The refinement involves differentiating between the intrinsic and variable fictional dimensions of land as commodity. The former refers to the fundamental characteristics of land that distinguish it from genuine commodities. For example, land cannot be manufactured to respond to increases in demand. The latter relates to mutable dimensions-meanings, values, functions-that differ between social groups and classes and take distinct forms in different social and historical contexts. The intrinsic and variable fictional dimensions of land as commodity are simultaneously present in processes of land commodification and have a significant bearing on agrarian politics.
For instance, intrinsic characteristics compel capitalist states to create and sustain land markets, including eradicating or weakening alternative forms of exchange, which generates political conflict. Meanwhile, variable dimensions stoke political tensions as different social groups and classes attempt to impose their meanings and values of land on the state and law (Goodwin, 2017). The Ecuadorian case shows that these struggles take particular forms at different historical moments and political tensions connected to intrinsic and variable dimensions shift as capitalism evolves.
Unpacking the fictitiousness of land as commodity therefore encourages eschewing static, essentialized readings of land commodification and viewing it as a dynamic, contextual process. Viewing fictitious commodification from this perspective cautions against seeing the development of land markets as a linear or teleological process. Indeed, Karl Polanyi's claim that creating land markets was one of the 'weirdest of all the undertakings of our ancestors' might well become increasingly prophetic as the social and ecological limits of land markets are reached and current and future generations search for more sophisticated institutions to manage land and mediate human-nature relations.