The evolution of communal land use provides vital insight into diverse processes of urbanization and industrialization in Mexico. The ejido, paragon of communal land tenure in Mexico during the twentieth century, represents a particularly insightful subject of analysis for researchers. Recent decades’ scholarship on ejidal development reveals that the ejido’s economic, legal, and cultural position in Mexican society resulted in urbanization patterns that often deviated from non-ejidal lands. The northern Mexican borderlands are illustrative of this phenomenon. In the state of Chihuahua during and after the Mexican Revolution (1910–1917), nationalized lands and natural resources were redistributed by the federal government in the form of inalienable agrarian grants. The ejido system soon came to dominate Chihuahuan and other state landscapes (Figure 1). As the century progressed, the intrinsically rural ejido increasingly abutted or was incorporated into burgeoning urban spaces, becoming “the most important source of land for development in Mexican cities” (Lombard Reference Lombard2016, 2703). In Ciudad Juárez, Chihuahua’s most populated city, located on the US-Mexico border, ejidal development was influenced by political, economic, and spatial factors particular to its borderlands geography. The distribution of industry and investment in the city’s formerly peripheral ejidal lands reveals both the complexity of ejidal urbanization processes in Mexico and how changes to communal land use shaped incipient borderlands industrialization.
Figure 1. Ejidos of Chihuahua. Sources: Open Street Map; Natural Earth Data; Registro Agrario Nacional. Credit: Erik Steiner.
Recent technologies provide scholars with novel tools to analyze established research subjects anew, and ejidal urbanization is no exception. In particular, the use of visual data and artificial intelligence allows us to construct innovative arguments that complicate our existing understanding of periurban and urban built environments, arguments formerly obscured by conventional sources. Ciudad Juárez, inhabiting vital borderlands space, represents an ideal case study. This article utilizes Google Street View (GSV) imagery beginning in 2007 and Geographic Information System (GIS) satellite imagery originating in the 1970s to provide visual evidence of industrialization and investment trends in Juárez, trends whose origins can be traced via conventional sources to postwar decades. Some of the conventional sources we use to historicize and analyze the GSV and GIS imagery are Mexico’s official state journal, borderlands newspapers, and Juárez census data. Our methodology reveals how ejidal urbanization in borderlands Juárez diverged from ejidal urbanization patterns in interior Mexican cities.
The key to our multidisciplinary methodology is its use of machine learning to read tens of thousands of GSV images. A diverse team of researchers and research assistants examined thousands of images to identify visual markers of capital investment. Markers included such factors as the conditions of buildings and public infrastructure (e.g., paved or dirt roads). The team then selected more than eight hundred GSV images that they agreed conformed to the following three categories: “invested,” “neutral,” and “impoverished.” This subset of images was fed into a state-of-the-art computer vision model for training, a process in which the model learns to recognize visual patterns that exist in a set of images. The model then built “invested,” “neutral,” and “impoverished” classes that could be compared to new sets of GSV images. The trained model, when fed a new GSV image, returned a measure of confidence (expressed in percentages) that “tagged” the image to a certain category. In other words, the model uses visual patterns it recognizes from the images selected by our team via close reading to classify new images, approximating how close they are to our examples (for a broader discussion of the process and evaluation of results using this computer vision model, see Barleta et al. Reference Barleta, Carrillo, Frank and Steiner2020).
Conscious of the method’s limitations (as well as the complexity of the subject), our research team adopted several conservative measures to ensure the integrity of our analysis. First, we kept classes simple and did not use the model to identify “nuances.” Second, we visually inspected samples of the tags predicted by the artificial intelligence model (which is standard practice in the field.) Third, we discarded all images with tags lacking very high confidence scores (greater than 85 percent). Fourth, we analyzed the resulting data by identifying large patterns that were difficult to detect in single observations. The fact that the resulting patterns emerged when we plotted the points on a map, and are consistent with the literature on the subject, is a strong indication of the reliability of the results. Finally, the data set was analyzed in conjunction with archival and historical sources that allowed us to corroborate patterns and, most importantly, propose an interpretation. We derived no causal explanation from the quantitative methods alone.
Our innovative methodology is vital to exposing previously unseen patterns of development in Juárez as well as to complicating the study of ejidal urbanization. In general, current scholarship presents the urbanization of Mexico’s ejidal lands in the following typical pattern: ejidos on the periphery of urban areas (periurban) were often sites of informal settlement, rapid and chaotic urbanization, and a lack of infrastructure and investment in the decades following Mexico’s 1960 urban turn (Salazar, Reis, and Varley Reference Salazar, Reis, Varley, Reis and Lukas2022; Barleta et al. Reference Barleta, Carrillo, Frank and Steiner2020; Schumacher et al. Reference Schumacher, Durán-Díaz, Gutiérrez-Juárez and 'González-Rivas2019; Lombard Reference Lombard2016; Parramond Reference Parramond2008; Jones and Ward Reference Jones and Ward1998). Following this classic model, inalienable yet invaded ejidal lands were increasingly expropriated by the state beginning in the 1970s to regularize existing migrant settlements and facilitate new public works projects (for more on the history of the ejido and land tenure regularization in Mexico in the latter half of the twentieth century, see Appendini Reference Appendini2001; Azuela and Duhau Reference Azuela, Duhau, Fernandes and Varley1998; Durand Reference Durand1983). Often, this resulted in the creation of periurban landscapes characterized by impoverishment and sprawl. This informal urbanization process, sanctioned by a corporatist state determined to maintain economic growth and social order, circumvented the ejido’s protected legal status ((Lombard Reference Lombard2016, 2704–2705). Changes to ejidal lands in Juárez, however, deviated from this formula in specific ways. After 1960, local, regional, and transnational elites primarily drove the urbanization of the city’s ejidal lands, not rural-to-urban migrants in informal settlements. These actors forged a modernization regime that privatized periurban ejidal lands and waters for the principal benefit of private industrial parks and foreign-owned factories (maquiladoras).Footnote 1 The result: current and former ejidal lands within Juárez’s urban boundaries often demonstrate high levels of investment, infrastructure, and planning, outcomes that diverge from the classic ejidal urbanization model in Mexico.
This article, as a result, explores the origins of private industrial parks in communal agrarian lands and the distribution of investment in Juárez by combining spatial and historical methodologies. For this particular study, the city’s residential street network was sampled to obtain twenty thousand random points with a minimum distance of twenty-five meters between points. A GSV image was downloaded for each point, but given GSV limitations, only 11,620 points were matched with a street-level image. Images were then processed via our computer vision model that tagged images into “invested,” “impoverished,” or “neutral” categories. Only GSV images tagged with greater than or equal to 85 percent confidence were considered for the current analysis.
The Ejido and the City
The ejido was one of the “most extraordinary social compromises” to emerge from the Mexican Revolution (Ginzberg Reference Ginzberg2019, 552). This modern iteration of communal land tenure, modeled on indigenous customs, redistributed nationalized natural resources to rural peasants beginning in the 1910s. Born of radical land reform ideology and enshrined in Mexico’s 1917 Constitution, the ejido conferred usage rights, not ownership, to its members (ejidatarios). Despite its social objectives and radical pedigree, the ejido system often reinforced ejidatarios’ cultural and economic differences (Rosemblatt Reference Rosemblatt2018). The redistribution of nationalized lands and waters via ejidal grants intensified during President Lázaro Cárdenas’s term (1934–1940). New federal bureaucracies administered these grants as well as agrarian credit, transforming ejidatarios into clients of the state. Cárdenas, under pressure from rising population rates and rightist movements, abandoned his more radical, domestic-focused agenda by the end of his administration. In doing so, he inaugurated Mexico’s conservative turn and ensuing embrace of Mexican-US cooperation, economic expansion, and rural modernization.
A major consequence of Mexico’s 1940 conservative turn was the escalation of ejidal land invasion by private interests, migrants, landless peasants, and fellow ejidatarios. Legal restrictions proscribing the sale of national lands and waters and the private consolidation of land were eased following World War II (Aboites Reference Aboites Aguilar2019, 1169). After Mexico’s 1960 urban turn, the vulnerability of periurban communal lands made them favored targets for migrant settlers and new public works projects, leading to a rise in ejidal expropriations in the 1970s. A series of constitutional reforms enacted by President Carlos Salinas beginning in 1991 legalized the sale and rental of ejidal lands, formalizing processes that had occurred unlawfully throughout Mexico for decades (Jones and Ward Reference Jones and Ward1998, 77–78).
Ejidal urbanization played a pivotal role in the development of Ciudad Juárez, a northern borderland metropolis with a rich agricultural past. After the Mexican-American War ended in 1848, the Mexican settlement of El Paso del Norte was divided by a new international border limited by the Rio Grande (or Río Bravo in Mexico) in the north and east and the mountainous desert to the city’s west and south. The city, inhabiting vital geographical, political, and economic space, served as a main hub for US-Mexican trade, transportation, and migration beginning in the late nineteenth century (García Reference García1981). In 1888 the city was renamed in honor of the former Mexican president and historical benefactor of local agrarian communities Benito Juárez. In fact, President Juárez’s 1860s grant of irrigable land to the agricultural community of Senecú, to the east of El Paso del Norte, was key to the city’s expansion. It effectively created an interstitial space between town and country that oriented future urban development eastward and along the river (Santiago Quijada Reference Santiago Quijada2011, 71–72, 121). As Figure 2 demonstrates, the pueblo of Senecú abutted the river and international border to the east of Juárez’s urban core, occupying riparian lands used largely for growing cotton.
Figure 2. Undated, early-twentieth-century map. Credit: Mapoteca Manuel Orozco y Berra (accessed at https://linealist.wordpress.com/2016/09/19/paso-del-norte/).
The rapid growth of industrialization, population, and migration in Mexico after 1940 inaugurated a dynamic era of urbanization in Juárez. In the 1960s the federal government initiated two pivotal programs: the Programa Nacional Fronterizo (National Border Program, PRONAF) in 1961 and the Border Industrialization Program (BIP) in 1965. These projects sought to increase economic activity in Mexico’s northern borderlands. PRONAF developed infrastructure and services near ports of entry to attract US investors and tourists. US investment, often in the tourist sector, had been a transformative economic and cultural force in Mexico since the early twentieth century (Velázquez García and Balslev Clausen Reference Velázquez, Alberto and Balslev Clausen2020). BIP, limited to the northern Mexican borderlands, created a free trade zone that enticed foreign manufacturers to relocate to borderlands industrial parks (Mize Reference Mize2008, 143). BIP sought to mitigate the termination of the Bracero Program (1942–1964), a series of guest labor accords that encouraged millions of rural Mexicans to migrate north to the United States. Former braceros amassed at Mexico’s northern border would, in theory, acquire jobs in domestic infrastructure projects and US-owned maquilas (de la O Reference de la O2007, 32). Contrary to BIP and PRONAF’s original intent, however, Juárez’s maquila workforce was soon dominated by women, reflecting the burgeoning national trends of domestic female migration, the feminization of domestic low-wage labor, and mass undocumented male labor migration to the United States after 1965 (Durand Reference Durand1994, 237–238).
At first glance, the urbanization of Juárez’s ejidal lands in the northern borderlands reflects similar trends observed in other major cities in Mexico’s interior. Urbanization and land development in Juárez were politicized processes influenced more by the partisan and economic interests of dominant political groups than the technical policies of public planners (Llera Pacheco Reference Llera Pacheco2000, 64–65). Given the importance of cities in general—and borderlands cities in particular—to Mexico’s post-1940 economic growth, both land use and urban development in Juárez and other major metro areas reflected the priorities of the party that dominated politics nationally until the year 2000, the Institutional Revolutionary Party (PRI). Consequently, the development of Juárez’s main periurban ejidos—Senecú, Zaragoza, and Salvárcar—reflected national trends. Located in modern Juárez’s eastern sector, all in proximity to the Rio Grande, these ejidos occupied the former liminal space on the eastern outskirts of the city’s historic urban core.Footnote 2
The evolution of the Juárez ejido of Zaragoza, for example, parallels trends documented in periurban ejidos in Guadalajara, Mexico’s third-largest metropolitan area and a classic model of informal ejidal urbanization (note the location of Zaragoza in Figure 2). Like the periurban ejido Santa Maria Tepetitlán and others in Guadalajara, Zaragoza’s initial ejido petition occurred in the 1920s, with an augmentation of ejidal hectarage taking place around midcentury. Shortly thereafter, an initial period of regularization of informal settlement through federal expropriation began in the 1970s, followed by substantial deregulation beginning in the early 1990s resulting from Salinas’s reforms (Barleta et al. Reference Barleta, Carrillo, Frank and Steiner2020; Jones and Ward Reference Jones and Ward1998).Footnote 3 As early as 1976 and through to the 1990s, then, the expropriation of ostensibly agrarian and inalienable ejidal lands occurred in Juárez, as in Guadalajara, to assist in the city’s “regular and planned growth.”Footnote 4 The state’s rationale of regularization was applied nationwide for periurban ejidal expropriations, not only in Juárez and Guadalajara (western Mexico) but also in major cities like Santiago de Querétaro (central Mexico) (Figure 3).
Figure 3. Regional overview map. Sources: Open Street Map; Natural Earth Data. Credit: Erik Steiner.
Similar to Juárez and Guadalajara, ejidal lands in the city of Querétaro experienced a dramatic change in the latter half of the twentieth century. In 1970 agriculture dominated both the economy and land tenure regime of the municipality of Querétaro.Footnote 5 However, informal settlements on the city’s periphery, established chiefly on periurban ejidal lands, proliferated beginning in the 1970s, a trend that directly informed the city’s urbanization and expansion of its city limits (Jorge Chavarría Bravo and Martha Sánchez A, qtd. in Icazuriaga Montes Reference Chavarría Bravo and Sánchez A.1994, 447). In the ensuing decades, Querétaro’s agricultural and ejidal lands were gradually subdivided for private and public housing due, in large part, to the intensification of rural-to-urban migration and industrialization. From 1960 to 1970, industrial parks were introduced into the municipality. After 1990s reforms the expropriation of lands in periurban ejidos like Santa Rosa Jáuregui was driven by urban public works projects, the regularization of existing informal settlements, and public municipal developments like sports fields and parks.Footnote 6
While aspects of the evolution of ejidal land use in Juárez paralleled patterns in interior Mexican cities like Guadalajara and Querétaro, the urbanization of communal lands in this key borderlands city diverged in crucial ways. Visual analyses of periurban ejidal landscapes in Juárez confirm this. GSV imagery illustrates that former and current ejidal lands within Juárez’s urban limits often demonstrate high levels of investment, infrastructure, and planning. Figure 4 is a random sample of an image taken from a residential location in the city’s formerly periurban, agrarian eastern sector near the former ejidal lands of Zaragoza.Footnote 7 Figure 5 represents another random sample from the city’s eastern sector, a public intersection located on lands comprising or abutting the former ejido of Senecú. These images represent examples tagged as “invested” by our computer vision model: paved roads and sidewalks, the presence of utilities, and appearance of formal planning. Conversely, Figures 6 and 7 depict urban landscapes the model identified as “impoverished”: unpaved or crumbling roads and/or sidewalks, vacant lots, graffiti, informality, and/or a lack of privacy. These random samples were taken from non-ejidal lands in Juárez’s western sector, namely the city’s historic center (Figure 6) and current southwest (Figure 7). The GSV images depicted in Figures 4 through 7 date from 2010 to 2019.
Figure 4. GSV image, 2010–2019. Eastern Ciudad Juárez. Source: Google Street View.
Figure 5. GSV image, 2010–2019. Eastern Ciudad Juárez. Source: Google Street View.
Figure 6. GSV image, 2010–2019. Ciudad Juárez city center. Source: Google Street View.
Figure 7. GSV image, 2010–2019. Southwestern Ciudad Juárez. Source: Google Street View.
In the case of periurban and urban ejidal lands in Guadalajara, GSV images often display the opposite scenario of Juárez’s former ejidal lands: distressed landscapes, a lack of investment and infrastructure, haphazard planning, and impoverishment. Figures 8 and 9 show sample GSV images tagged as impoverished from Guadalajara. In the case of Querétaro, after 1960, periurban ejidal lands were often sites of regularized informal housing, public works projects, and public spaces. Figures 10 and 11 show sample GSV images from former ejidal lands in Querétaro. Ultimately, national and municipal data reveal that significant portions of Juárez’s former and current ejidal lands, unlike those in interior Mexican metropolitan areas, are composed of or adjacent to private industrial parks containing multinational maquiladoras, as well as private subdivisions featuring high levels of investment and income, and fewer public spaces (IMIP 2014). GSV and GIS data and conventional sources demonstrate that ejidal lands in Juárez were privatized in a more systematic and quasi-official process, one that disproportionately benefited local elites, private US industry, and private residential subdivisions.
Figure 8. GSV image, 2010–2019. Southeastern Guadalajara. Source: Google Street View.
Figure 9. GSV image, 2010–2019. Southeastern Guadalajara. Source: Google Street View.
Figure 10. GSV image, 2016. Northern Querétaro. Source: Google Street View.
Figure 11. GSV image, 2016. Northern Querétaro. Source: Google Street View.
The process of ejidal urbanization in Juárez deviated from other examples observed in Mexico due largely to the city’s political, economic, and spatial geographies. With development dictated by mayors whose power resided in city councils dominated by a single party (often the PRI), land investors formed networks with politicians and not urban planners (Llera Pacheco Reference Llera Pacheco2000, 65). Moreover, local political bosses, land developers, and businessmen were often the same people, with close ties to Mexican heads of state and US industrialists (Vázquez Ruiz Reference Vázquez Ruiz2004, 106). A regime of development via private initiative and binational free trade—encouraged by the state and predicated, in part, on the privatization of ejidal lands and waters—emerged in Juárez. Located on the literal and ideological front lines of globalization, Juárez was at the forefront of the privatization of public resources and functions. This process resulted in the rapid growth of private industrial parks and maquiladoras on former ejidal lands and shaped the distribution of investment and inequality in Juárez.
Ejido Senecú and the Antonio J. Bermúdez Industrial Park
The pivotal role of communal lands in Juárez’s urbanization and industrialization is embodied by Senecú, an ancient settlement and postrevolutionary ejido that no longer exists. A community that occupied the Valley of Juárez for centuries, Senecú was designated an ejido in 1938 (Bowden Reference Bowden1971, 129).Footnote 8 It was located in Juárez’s eastern sector along the Rio Grande, on whose waters it depended (Figure 12). Senecú was the ejido nearest to Juárez’s urban core, and its rural produce was essential to the city’s growth (Santiago Quijada Reference Santiago Quijada2011, 78). Despite this past, Senecú became the first ejido in Mexico to wholly privatize after Salinas’s 1992 reforms.Footnote 9 Thus, the evolution of communal land tenure in Senecú, as well as the ejido’s ultimate dissolution, is revelatory. The persons most responsible for the establishment of the regime that led to Senecú’s dissolution were local PRI politicians, landowners, and businessmen Antonio J. Bermúdez and his nephew Jaime Bermúdez Cuarón.
Figure 12. Ejido Senecú, mid-twentieth century. Sources: Open Street Map; rough boundaries of the ejido obtained from Antonio Rebolledo, “Ejidos, una especie en extinción aquí,” El Diario [Juárez], June 2, 2013. Credit: Erik Steiner.
The rise of Antonio J. Bermúdez’s political and economic fortunes paralleled those of Juárez itself. Originally from the city of Chihuahua, Bermúdez moved to Juárez during the Revolution.Footnote 10 He amassed his wealth smuggling liquor across the border during US Prohibition. He cemented his position among the city’s social and political elite by marrying into the pre-Revolutionary landowning class and purchasing part of the infamous Terrazas hacienda (Wasserman Reference Wasserman1993, 113). In 1942 and 1943 Bermúdez served as Juárez’s municipal president, a position he accepted only at the behest of his “dear friend” President Manuel Ávila Camacho. Bermúdez’s influence in the PRI resulted in political appointments at the state and federal levels, ambassadorships to Middle Eastern countries, and the directorship of PEMEX, the state-run petroleum company. By the time Bermúdez was named PRONAF director in 1960, he was intimately familiar with the city’s physical and political geographies as well as the PRI’s industrialization and globalization goals. Jaime Bermúdez Cuarón would leverage his uncle’s extensive financial and political capital by establishing Mexico’s first private industrial park, in part, on Senecú’s ejidal lands.
The creation of the Antonio J. Bermúdez Industrial Park (AJB) in 1968 was a watershed moment in the development of Juárez and evolution of its ejidos. Jaime Bermúdez, one of twelve local businessmen who founded the AJB, was born into a prominent family that owned a cotton farm in the Valley of Juárez. Jaime saw a tremendous opportunity to increase the value of his family’s holdings when his uncle was named PRONAF director. In fact, he and other local business leaders lobbied the elder Bermúdez to support an industrialization study that proved instrumental to the passage of BIP.Footnote 11 The implementation of BIP and new tax structures provided crucial incentives for US companies to relocate to Juárez. And, as a US-educated engineer, Jaime had connections to US business and industry. He became a leading BIP booster, traveling across the United States to attract companies to Juárez. Jaime’s first US client arrived in 1967, and the following year, Jaime broke ground on the city’s (and nation’s) first private industrial park.Footnote 12 Named after Antonio, this maquila industrial park was promoted as the development model not only for Juárez but other borderlands cities.Footnote 13 Yet the AJB occupied lands and used waters that legally belonged to the ejido of Senecú and, therefore, the Mexican people (Figure 13) (Santiago Quijada Reference Santiago Quijada2011, 415–416).
Figure 13. Ejido Senecú and Antonio J. Bermúdez Industrial Park. Sources: Open Street Map; Registro Agrario Nacional; Instituto Municipal de Investigación y Planeación Ciudad Juárez (IMIP); Rebolledo, “Ejidos, una especie en extinción aquí.” Credit: Erik Steiner.
Although the AJB was created in 1968, its foundations traced to the 1940s. The adoption of import substitution industrialization (ISI) policies affected northern Mexico’s ejidos more than most. Many were vast, industrialized farms growing commercial crops. ISI increased demand for domestic raw materials, intensifying cotton production in the Valley of Juárez (Bustillos Durán Reference Bustillos Durán2004, 229). As the PRI pursued economic autonomy via ISI, it also sought to integrate Mexico’s economy into global markets and erase international borders (Carrillo Reference Carrillo2019, 61). Chief among the priísta proponents of globalization was Carlos Lazo, a prominent architect and technocrat whose modernism was heavily influenced by trips to the northern borderlands. In an influential 1950 article titled “Aching Border,” Lazo decried vice and migration in the borderlands, attributing these social ills to the region’s lack of infrastructure.Footnote 14 He called on the nation’s planners and engineers to develop the borderlands’ built environment and integrate it with that of the southern United States. Lazo’s vision, as well as his role as federal secretary of Communications and Public Works from 1952 to 1955, planted the seed for PRONAF, BIP, and private industrial parks in Juárez.
By 1950 Juárez was more than a key borderlands city undergoing rapid urbanization. It was a vital proving ground for Mexican modernism due to the dynamic political, economic, and physical space it inhabited. Small farms, especially communal and subsistence ones, did not fit into this emerging globalized model of Mexican industrialism and free trade. The 138-hectare Senecú did not produce crops on the scale of the large industrial ejidos in other parts of the state or region. Plus, the physical space inhabited by ejidos like Senecú—situated between an international border and the boundaries of a growing metropolis—made them prime locations for US-focused trade in the eyes of Juárez’s industrialists (Vázquez Ruiz Reference Vázquez Ruiz2004, 100). These impresarios knew that industry required infrastructure, land, and water and that arid northern Chihuahua possessed little water or infrastructure. Juárez’s periurban ejidos, with their legally tenuous public lands and waters, represented ideal candidates for private industrial development.
The mid-1950s were a turning point in the urbanization of the Valley of Juárez. The process of urbanizing and privatizing Senecú’s public resources originated in 1955 with the construction of the upscale neighborhood subdivision “Club Campestre de Ciudad Juárez,” a private country club located on Senecú’s western margins (Santiago Quijada Reference Santiago Quijada2011, 366–367). The Club Campestre project led Juárez planners to expand the road network that connected the city’s historic core with its eastern hinterland. In late 1955 plans to pave nearly four hundred kilometers of road were finalized.Footnote 15 Months later city officials announced that Vicente Guerrero, a major Juárez avenue, would be extended to the east.Footnote 16 Figure 14, which uses Juárez’s modern street grid, illustrates Vincente Guerrero’s location and role in the historical expansion of the city’s boundaries.
Figure 14. Growth of Juárez. Sources: Open Street Map; Registro Agrario Nacional; IMIP; the Atlas of Urban Expansion. Credit: Erik Steiner.
Juárez’s eastward growth at midcentury, symbolized by the construction of private subdivisions and underwritten by public resources and infrastructure investment, revealed an incipient process of urbanization in periurban ejidal lands that was not haphazard. In fact, urban planners in Juárez, under the guidance and patronage of local officials and businessmen like the Bermúdez family, followed official regulatory plans proposed in 1958 and 1962 (Bermúdez Reference Bermúdez1966). These official plans’ overarching objective was formal, planned growth toward the agricultural and riparian eastern regions of the city (Rodriguez and Rivero Reference Rodriguez and Rivero2011). With PRONAF and BIP encouraging rapid industrialization and free trade by the mid-1960s, the city’s economic and political elites rushed to acquire eastern rural land for development, speeding urbanization.Footnote 17 Once land was acquired, local, state, and national politicians, many investors themselves, provided generous incentives and resources (especially water of “good quality”) to private manufacturers and industrial park owners.Footnote 18
The privatization of public resources in 1960s Juárez was facilitated by the privatization of public functions. In particular, municipal road building represented a lucrative windfall for private local businesses. One such firm, Constructora y Urbanizadora Nacional SA (CUNSA), was owned by PRONAF chief Antonio J. Bermúdez himself. CUNSA was awarded government contracts for road building in the 1950s and 1960s. In fact, the PRONAF head’s business consortium, Grupo Bermúdez, was involved not only in construction but also in Juárez’s commerce and service industries (Vázquez Ruiz Reference Vázquez Ruiz2004, 106). Bermúdez’s self-dealings did not go unchallenged, however. Mexico’s National Chamber of Commerce accused the Juárez Chambers of Commerce and Industry of corruption in 1963.Footnote 19 (Despite this, CUNSA received another road contract five years later).Footnote 20 At the same time, Juárez’s rural sector denounced the expansion of the city’s urban boundaries. Local agriculturalists argued that the elites’ rezoning of rural lands to urban lands was being done inequitably and to their detriment.Footnote 21
The establishment of the privately owned AJB in part on communal lands represented a key inflection point in Juárez’s urbanization. Following the park’s opening in 1968, more foreign companies relocated to the city and the maquiladora industry thrived. Five new maquilas were added to the AJB in its first two years. In 1972, 125 million pesos were invested in a park expansion that sought to accommodate thirty additional maquilas.Footnote 22 The parallel growth of the AJB and the maquila sector in Juárez continued throughout the 1970s. The peso’s devaluation in the 1980s made cheap Mexican labor more attractive to US companies, further galvanizing the maquila industry. By 1985, Juárez was home to seven industrial parks. Between 1982 and 1987, the number of maquilas in the city doubled, and the number of employees tripled.Footnote 23 BIP was a resounding success for Juárez as well as Mexican modernization and globalization. Yet not all residents profited from Juárez’s regime of industrialization via maquila, ejidatarios in particular.
While Juárez’s private sector reaped the benefits of industrial and economic growth, periurban ejidos faced mounting environmental and resource challenges. Water, vital to both farm and maquila, was distributed more inequitably after 1965. Beginning in the mid-1960s, city officials, often private investors themselves, created wells to ensure maquilas’ access to water.Footnote 24 Ejidos, administered by federal agencies, were not granted similar assistance. In fact, postwar urbanization and privatization trends saw agriculturalists throughout the northern borderlands compete with cities in zero-sum contests for water that encouraged the rapid proliferation of wells and pumps extracting groundwater (Walsh Reference Walsh2011, 55). As water in Juárez transitioned in earnest from a public to private (and rural to urban) resource by the late 1970s, expanded groundwater pumping increased the salinity of the irrigation water ejidos depended on—a problem that had steadily worsened over preceding decades.Footnote 25 This issue affected small farms predominantly, as large landholders were able to maintain or increase crop production due to their disproportionate access to state-controlled resources.Footnote 26 Moreover, new private subdivisions exacerbated ejidos’ water woes. By the late 1980s, for example, some neighborhoods in Senecú’s periphery had “converted ditches into dumps,” obstructing the passage of fluids and contaminating the water supply of Senecú as well as other rural and agricultural communities in the area.Footnote 27
As environmental ills and resource inequality increasingly threatened ejidatarios’ lands and livelihoods, Juárez struggled to attract seasonal farm labor due to mass migration to the United States. This phenomenon only intensified after the passage of BIP. Rural male laborers, accustomed to working in large numbers in US agriculture and industry since World War II, preferred US jobs due to their relatively higher wages. In 1978, for example, the director of the Regional Union of Ejidatarios reported that dozens of men from the interior were arriving daily in Juárez with the goal of crossing into the United States despite labor shortages in the city.Footnote 28 This trend continued into the 1980s, as male migrants used Juárez primarily as a temporary stopover on their US labor journeys.Footnote 29 And although maquila jobs were an option in theory for the largely male pool of labor migrants, maquila employers preferred to hire young, single, and childless women who were perceived to be more “compliant” (Minian Reference Minian2018, 27–28).
The urbanization of Juárez’s ejidal lands via a regime that privatized public resources was sanctioned by Mexican officials well before the country’s “neoliberal turn” in the late 1980s and 1990s. One particular example from the early 1980s makes this clear. Agriculturalists from the Valley of Juárez accused Jaime Bermúdez of invading eleven thousand hectares of their land. They claimed Bermúdez illegally possessed these lands since the 1960s with the consent of Chihuahua’s then governor, the priísta Óscar Flores Sánchez.Footnote 30 In 1976, Flores Sánchez became Mexico’s attorney general, the official most responsible for enforcing federal laws at a critical legal juncture for the ejido system. Jaime’s illicit actions, along with his uncle’s self-dealings and the complicity of state and federal officials, exposed the economic boon and legal impunity of privatizing ejidal lands, especially in the northern borderlands. These lessons were not lost on the US-educated technocrats who dominated Mexican politics nationally beginning in the 1980s.
President Salinas’s 1992 ejidal deregulation merely formalized processes that had occurred throughout Mexico for decades. By constructing Juárez’s (and Mexico’s) first private industrial park on lands belonging to and abutting the ejido of Senecú, officials and investors helped erode the legal and cultural barriers to privatizing ejidal lands (Santiago Quijada Reference Santiago Quijada2011, 567). Within two years of Salinas’s reforms, Senecú’s remaining ejidatarios were granted full individual domain of its parcels. Six years later, Senecú’s members began the process to formally dissolve the ejido. And nearly a decade after early 1990s reforms, Senecú was fully privatized, marking another national first for Ciudad Juárez: the complete dissolution of a postrevolutionary ejido.Footnote 31
Senecú’s transition from a periurban to urban landscape and a public to private natural resource characterized by industrial parks and foreign-owned maquiladoras underscores the key divergences between ejidal urbanization in Mexico’s northern borderlands and its interior. This process also symbolized the broader project of Mexican modernization after 1940, one that prioritized rapid industrialization and globalization over social and resource parity. Crucially, the urbanization of Juárez’s eastern periphery—predicated, in part, on ejidal lands like Senecú’s as well as segments of ejidos Zaragoza and Salvárcar—shaped the spatial distribution of investment in the city after 1960. And despite Carlos Lazo’s midcentury modernist vision for Mexico’s northern borderlands, investment and infrastructure patterns in Juárez intensified inequality, a trend that is vividly demonstrated by GSV and GIS imagery and our computer vision model.
Spatial distribution of investment in Juárez: Comparisons and consequences
GSV and GIS imagery demonstrate that the urbanization of rural and communal lands in Juárez’s former eastern periphery frequently produced landscapes characterized by high levels of investment, infrastructure, and formal planning, a departure from the classic model of post-1960 ejidal urbanization in Mexico identified by past scholarship. As Figure 15 illustrates, the distribution of GSV images tagged as impoverished largely splits Juárez on a north-south axis between its more invested eastern half, home to the city’s principal (former and current) periurban ejidal lands, original private industrial parks, and private housing subdivisions, and its less invested western half, home to impoverished urban landscapes and informal housing dominated by the city’s historic core.
Figure 15. Distribution of Investment. Sources: Open Street Map; Registro Agrario Nacional; IMIP; Google Street View. Credit: Erik Steiner.
The high level of investment and planning encountered in former periurban ejidal lands in Juárez diverges from trends witnessed in Guadalajara, the nation’s third-largest metropolitan region. GSV images from 2010 to 2019 and GIS data provide a revealing portrait of the distribution and concentration of tags characterized as impoverished in both cities. Figure 16 demonstrates that impoverished, informal landscapes dominate Juárez’s urban core, not periphery, as is seen in Guadalajara in Figure 17. In particular, the northwest is home to Juárez’s historic urban center and neighborhoods (colonias), not former or current ejidal lands. Income levels in the western sector also pale in comparison to Juárez’s more formally developed eastern sector (excluding the current city’s extreme south and southeast peripheries), which is home to private maquila industrial parks that occupy or abut the former and/or current periurban ejidal lands of Senecú, Salvárcar, and Zaragoza. Note the distribution of industrial parks and description of city sectors in Figure 18.
Figure 16. Income and Impoverishment. Sources: Open Street Map; Instituto Nacional de Estadística y Geografía 2010; AGEB Ciudad Juárez (https://purl.stanford.edu/bx274kk2004); Google Street View. Credit: Erik Steiner and Zephyr Frank.
Figure 17. Distribution of GSV Images Tagged as Impoverished, Guadalajara, Jalisco. Sources: Open Street Map; Natural Earth Data; Registro Agrario Nacional; Google Street View. Credit: Leonardo Barleta.
Figure 18. List of Juárez industrial parks and zones by urban sector. Source: IMIP, Actualización diagnóstica sociodemográfica y económica del programa de desarrollo urbano de Ciudad Juárez, Chih. (Juárez: IMIP, March 2014), 15.
The distribution of GSV images tagged as impoverished differs markedly in Guadalajara. In Figure 17, impoverished tags cluster exclusively on the urban periphery, the location of the city’s former and current periurban communal and rural lands. The positions of these clusters suggest successive waves of invasion of ejidal lands by migrants seeking housing on peripheral and vulnerable lands deficient in infrastructure and planning, a process that reflects Mexico’s classic ejidal urbanization model. Guadalajara’s historic (and geographic) urban center and surrounding older colonias do not demonstrate the low levels of infrastructure and investment that characterize Juárez’s core. Plus, Guadalajara, located in the Atemajac Valley in western Mexico’s Sierra Madre Occidental, was not shaped spatially or politically by an international border serving as the historical entrepôt for foreign capital, trade, and industry.
As Figure 16 indicates, the evolution of ejidal lands and ensuing concentration of investment and wealth in Juárez’s eastern sector contributed to the concentration of informality and poverty in non-ejidal lands after 1960. Rather than clustering on the urban periphery like Guadalajara, informal settlements in Juárez grouped nearer the city’s core, occupying or abutting the Sierra de Juárez mountains and hills immediately west and southwest of the historic city center. Since ejidal and riparian lands in the Valley of Juárez were coveted by officials and investors for their manufacturing potential (i.e., access to cheap land and water), the lands most available to poor rural-to-urban migrants were located mainly in arid, mountainous terrain. These informal settlements were often populated by workers from the maquila industry, Juárez’s largest employer. Maquila industrial parks were located disproportionately in the current city’s east and center, as illustrated in Figure 18. Crucially, the maquila workforce was dominated by women by the 1990s (Wright Reference Wright2004, 373).
Gendered poverty and violence were major outcomes of post-1960 industrialization and globalization trends in Juárez (González Rodríguez Reference González Rodríguez2010; Bermúdez Urbina and Evangelista García Reference Marina2017). While the processes of ejidal urbanization did not contribute directly to violence against women, GSV and GIS imagery provide insight into the evolution of ejidal land use and ensuing distribution of investment, inequality, and violence in Juárez. In particular, significant correlations emerge when comparing investment and femicide trends in the city. Because femicide is one of the most tangible manifestations of inequality and violence in late twentieth-century Juárez (along with narcoviolence)—and an act intimately associated with maquila work, poverty, public space, and a lack of investment and infrastructure—it is worth briefly exploring the linkages between investment and femicide patterns here.
Municipal data and GIS imagery demonstrate that gendered poverty and violence were concentrated largely—but not exclusively—to Juárez’s western sector (Cervera Gómez and Monárrez Fragoso Reference Cervera Gómez2010, 14). In fact, by the early 2000s, six of the city’s top seven “key femicide zones” were located in Juárez’s northwest and center-west (Cervera Gómez and Monárrez Fragoso Reference Cervera Gómez2010, 60–61, 66). Figure 19 (taken from a municipal study of femicide trends from 1993 to 2010) reveals that the urban spaces where female victims’ bodies were discovered map closely onto the distributions of impoverishment and income shown in Figure 16. Western Juárez’s non-ejidal lands provided the bulk of informal housing for migrants and maquila employees, housing often constructed of cardboard or wood pallets.Footnote 32 As Figures 16 and 19 illustrate, these urban landscapes of poverty and violence were situated far from the private industrial parks, maquilas, and subdivisions dominating the city’s east. Perhaps the most visible indicator of Juárez’s unequal investment regime was the presence or lack of infrastructure (Grineski and Collins Reference Grineski and Collins2008, 261).
Figure 19. Spatial density of femicide in Ciudad Juárez by location of victim residence and crime scene, 1993–2010. Source: Map by Luis Cervera Gómez, in Luis Cervera Gómez and Julia Monárrez Fragoso, Sistema de información geográfica de la violencia en el municipio de Juárez, Chihuahua: Geo-referenciación y su comportamiento espacial en el contexto urbano y rural, Reporte final (Juárez: Secretaría de Gobernación, 2010), 59.
Access to water and its infrastructure played a principal role in the spatial distribution of invested and impoverished landscapes in Juárez. In the 1950s, municipal projects that sought to expand water and sewer infrastructure incorporated older communities abutting Juárez’s historic center.Footnote 33 And colonias on the more immediate western and southern limits of the urban core were included in 1960s water studies.Footnote 34 However, by the early 1970s, shifting development priorities led to the rapid expansion of maquilas and industrial parks in Juárez’s east.Footnote 35 These private industrial parks and subdivisions eventually dominated public water resources and infrastructure investment. By the late 1970s, subdivisions in Juárez’s formerly rural, agricultural (and often ejidal) northeast were fully modernized.Footnote 36 In the meantime, the flow of migrants from the Mexican interior swelled Juárez’s informal settlements on lands lacking water and infrastructure, which were located largely in the western sector. By the 1990s, female migrants, most of them low-wage industrial laborers or domestic servants, had “overwhelmed” these settlements.Footnote 37
Road infrastructure, like that of water, reflected the duality of Juárez’s built environment. In particular, public transportation embodied the gendered inequities wrought by the city’s urbanization regime. Maquila workers, typically poor females, commuted long distances on foot from their colonias, often western slums lacking running water and basic services (González Rodríguez Reference González Rodríguez2010). Once in the city center they boarded company buses that delivered them to industrial parks in eastern and central Juárez. This daily trek, frequently undertaken alone and at night along dirt roads lacking streetlights and police, could be dangerous and deadly (Livingston Reference Livingston2004, 61).Footnote 38 During the 1990s, for example, local bus drivers were charged with using their vehicles and knowledge of the city’s less formal and secure road networks to prey on women traveling to and from work alone.Footnote 39 These serial murders were the most tangible manifestation of a stark correlation between the location of femicide crime scenes and Juárez’s roads (Wright Reference Wright2011, 713).Footnote 40 From 1993 to 2010 nearly a third of all femicide victims’ bodies were discovered on the city’s less invested and less secure western roadways (Cervera Gómez and Monárrez Fragoso Reference Cervera Gómez2010, 16).Footnote 41
GSV, GIS, and municipal data demonstrate that where a woman lived and died correlated robustly with Juárez’s disparate landscapes of investment and income. An “intimate connection” existed between a victim’s residence and the location where her body was discovered (Cervera Gómez and Monárrez Fragoso Reference Cervera Gómez2010, 14). And from 1993 to 2010, nearly half of all victims’ bodies were encountered in vacant lots (Cervera Gómez and Monárrez Fragoso Reference Cervera Gómez2010, 16). Where a woman lived and where her body was recovered were intrinsically linked to poor and divested, often public, spaces (Wright Reference Wright2004, 369).Footnote 42 As Figures 16 and 19 show, these factors converged regularly in western Juárez. Contrast these vacant lots, unpaved and unlit roads, public spaces, informal housing, and lack of utilities (our model’s classification of “impoverishment”) with Juárez’s more invested, planned, private, and generally wealthier eastern sector, home to the majority of the city’s maquilas. Researchers have observed that Juárez’s maquila industrial parks—featuring fewer public spaces, more robust planning and infrastructure, and ample access to water—were, by the 2000s, “more often located in formally developed areas inhabited by a more affluent populace” (Grineski and Collins Reference Grineski and Collins2008, 266).Footnote 43 These more private, secure, and wealthy landscapes were forged, in part, from former communal lands of Senecú as well as parcels from Zaragoza and Salvárcar.
While the causes of femicide are too multivariate and complex to fully explore here, our methodology provides a fresh perspective on the correlations between urbanization and violence trends in Juárez. Visual imagery reveals how the evolution of the city’s periurban ejidal lands contributed to distinct and disparate patterns of investment and income. This process informed the distribution of inequality, which forged two spatially and economically distinct cities by the 1980s—one more industrial, private, and invested in the east, the other more informal, public, and impoverished in the west. And in the 1980s and 1990s this tale of two cities took a grim turn as financial crises and neoliberal reforms intensified poverty and violence in Juárez (Chew Sánchez Reference Chew Sánchez2014, 267–68). Nearly two hundred women were officially identified as murder victims from 1993 to 1999 in Juárez (in 1999, the city was home to 330 mostly US-owned maquilas).Footnote 44 Nowhere in Mexico are the correlations between ejidal land use, industry, investment, and gendered inequality clearer than in Juárez.
Conclusion
Visual imagery, when analyzed via a computer vision model in combination with conventional historical sources, reveals a process of ejidal urbanization in the northern Mexican borderlands that diverged from the classic model observed in interior metropolitan regions of Mexico. Furthermore, our use of visual data and machine learning exposes the essential role of private industrial parks and multinational maquiladoras in the (d)evolution of communal land tenure in Juárez, a trend that began in the 1950s and reached its zenith in the 1990s. The process of ejidal urbanization, shaped by Juárez’s political, economic, and physical geographies, influenced the spatial distribution of investment and inequality in the city. This assertion is supported by GSV imagery sampled from various points throughout the city, as well as imagery from metro areas in interior Mexico. Regional variations in ejidal urbanization—particularly Juárez’s lower incidences of informal housing and impoverished landscapes in periurban and urban ejidal lands and higher incidences of inequality and poverty in non-ejidal and/or nonperipheral lands—suggest that further research is necessary to fully understand the evolution of ejidal land tenure and rural modernization in the US-Mexico borderlands.
Last, further analyses of ejidal urbanization trends in the northern Mexican borderlands will provide valuable perspective for scholars examining US-assisted development in Mexico during the twentieth century. The growing literature on the decentralization of Latin American states and deregulation of these states’ public resources and functions further emphasizes that Mexican neoliberal reform was less a period encompassing the late 1980s and 1990s and more a half-century-long process with roots in the 1930s and World War II (Offner Reference Offner2019; Fajardo Reference Fajardo2022). As one of the most celebrated social programs and radical commitments to resource sovereignty in Mexican history, the ejido remains a revelatory subject whose origin, (d)evolution, and ultimate fate continue to provide crucial insight into the global impacts of postwar modernization and globalization.
Acknowledgments
The authors would like to express their gratitude to Dr. Leonardo Barleta of the University of Nebraska Omaha for his generous assistance on earlier drafts of this article. They would also like to thank Dr. Zephyr Frank of Stanford University’s Department of History and Erik Steiner of the Center for Spatial and Textual Analysis at Stanford University for their contributions to this project.
Open access
