Hidden Farmworker Labor Camps in North Carolina: An Indicator of Structural Vulnerability

Abstract

Objectives. We used geographic information systems (GIS) to delineate whether farmworker labor camps were hidden and to determine whether hidden camps differed from visible camps in terms of physical and resident characteristics.

Methods. We collected data using observation, interview, and public domain GIS data for 180 farmworker labor camps in east central North Carolina. A hidden camp was defined as one that was at least 0.15 miles from an all-weather road or located behind natural or manufactured objects. Hidden camps were compared with visible camps in terms of physical and resident characteristics.

Results. More than one third (37.8%) of the farmworker labor camps were hidden. Hidden camps were significantly larger (42.7% vs 17.0% with 21 or more residents; P ≤ .001; and 29.4% vs 13.5% with 3 or more dwellings; P = .002) and were more likely to include barracks (50% vs 19.6%; P ≤ .001) than were visible camps.

Conclusions. Poor housing conditions in farmworker labor camps often go unnoticed because they are hidden in the rural landscape, increasing farmworker vulnerability. Policies that promote greater community engagement with farmworker labor camp residents to reduce structural vulnerability should be considered.

Agriculture is a major rural industry and among the most hazardous.^1–3 Although everyone who works in agriculture is exposed to substantial mechanical, chemical, and environmental hazards, migrant and seasonal farmworkers experience these hazards with the least reward. More than a million hired farmworkers are employed in US agriculture,⁴ approximately 150 000 of whom are employed in North Carolina alone.^5,6 Farmworkers are a vulnerable population. They are predominantly Latino immigrants and often do not have lawful documentation or speak adequate English; in addition, most have little formal education and limited incomes.⁷ Although working in a hazardous industry, farmworkers have limited access to health care.⁸ Political and economic forces that act on farmworkers increase their vulnerability by imposing physical or emotional suffering in patterned ways.⁹ The act of migrating is a substantial risk for many farmworkers.¹⁰ When faced with workplace hazards, they are often reluctant to complain for fear of losing a needed job.^11–13

Housing is an example of the structural vulnerability faced by migrant farmworkers. Structural vulnerability is a product of class-based economic exploitation and cultural, gender, sexual, and racial discrimination, as well as the complementary processes of depreciated subjectivity formation.⁹ Farmworker housing conditions are poor,^14–18 both in the local housing market^19–21 and in employer-provided farmworker labor camps.^22–24 Crowding, lack of access to sufficient bathing facilities, pest infestation, and structural damage are common to dwellings in farmworker labor camps.^22,25 Employer-provided farmworker housing seldom meets the requirements of state and federal regulations²² and is generally inhabited on a temporary basis for the period when the workers are employed, whereas local market housing is inhabited on a temporary or long-term basis depending on whether the residents migrate.

How farmworker housing is placed contributes to the persistence of substandard housing. Farmworker housing that is hidden from public view may receive less natural surveillance and become more vulnerable to crime,²⁶ including theft from workers,²⁷ theft of workers (human trafficking and slavery),⁵ and provision of substandard housing that does not meet regulations.²⁰ Hidden farmworker housing also limits access for individuals trying to provide services (e.g., clinic outreach workers) and advocacy groups (e.g., workers’ rights advocates) that can improve the conditions for farmworkers. Crime prevention through environmental design²⁶ argues that a built environment that enhances public visibility reduces crime. Natural surveillance enhances community accountability by using the built environment to enable passive community observation of events within a setting. If more of the public can observe farmworker housing, then the occupants will receive greater accountability from the public. These “many eyes” act as a deterrent for theft and vandalism and improve adherence to housing regulations. Inversely, hidden farmworker housing makes residents more vulnerable and makes the provider of substandard housing less accountable to the public.

This analysis used geographic information systems (GIS) to examine the location of farmworker labor camps. The aims were to delineate the proportion of hidden farmworker labor camps and to determine whether hidden camps differed from visible camps in terms of physical or resident characteristics.

METHODS

The data we collected were from a community-based participatory research (CBPR) project with the goal of describing the quality of farmworker housing and delineating the association of farmworker housing quality with health.²² Community partners included the North Carolina Farmworkers Project, Student Action with Farmworkers, and other clinics and organizations serving farmworkers in North Carolina. Data were collected from June through October 2010.

Sample

We recruited participants in 16 North Carolina counties in which large numbers of migrant farmworkers were employed: Caswell, Craven, Cumberland, Duplin, Edgecombe, Greene, Halifax, Harnett, Johnston, Lenoir, Nash, Person, Sampson, Wake, Wayne, and Wilson. We recruited “camps,” defined as temporary housing provided by employers, in each county based on lists provided by our community partners; we attempted to recruit any camps encountered during data collection that were not on these lists. We identified 226 camps. Residents in 36 camps declined to participate, the grower or contractor refused to permit participation in 4 additional camps, and residents of 186 camps agreed to participate, resulting in a camp participation rate of 82.3%. Camps that participated in the study were given a volleyball as a token of appreciation. No data were collected for camps that declined to participate.

Data Collection

Interviewers explained the study to camp residents and completed a census of the camp. Based on this census, they selected and obtained informed consent from 3 individual residents in each camp; each of these residents received a $30 incentive for participating. Data collection included interviews with 2 camp residents and a camp inspection completed with the assistance of the third resident. The interviews collected information about general housing characteristics, such as number of residents, number of dwellings, structure type, and housing conditions, and personal characteristics, such as age and health. We based the camp inspection on the housing quality standards promulgated by the North Carolina Department of Labor (NCDOL) but also included other items of interest to community partners and investigators.²²

We collected geographic data by making a Google Earth map in collaboration with community partners of known camp locations. This Keyhole Markup Language (KML) map layer of camp locations became the basis of the GIS for this analysis.²⁸ The coordinates from the KML of the camps were loaded into nüvi 260w GPS devices (Garmin Ltd., Olathe, KS) via a GPS Exchange Format. The GPS devices then served as navigational aids for data collectors who were conducting the housing study. This use of GPS was a means for “ground truthing” the camp locations that the community partners provided. Data collectors saved the coordinates of camps located during the course of data collection into the GPS devices. This method overcame challenges posed by geocoding addresses.²⁹ After we completed data collection, we created a final KML map of the study sites. The final map drew on the original KML map of camp locations plus corrections and additions from data collectors via their GPS devices. All camp location coordinates were verified; if verification was not possible, camps were excluded from the analysis (n = 6).

Measures

We included 3 measures of camp physical characteristics and 3 of resident characteristics. The number of camp residents had 4 categories: 1 to 6, 7 to 10, 11 to 20, and 21 or more. Number of housing units in the camp had the values 1, 2, and 3 or more. Housing type was the presence or absence of barracks in the camp. A barracks is a building built specifically to house farmworkers; it may be a freestanding structure or attached to an existing structure. Nonbarracks farmworker housing includes old farmhouses and trailers. Camps with barracks could also have nonbarracks housing, such as houses and trailers. Nonbarracks camps had only houses or trailers. H-2A status was a dichotomous measure that indicated whether any farmworkers with H-2A visas were living in the camp. An H-2A visa is a temporary agricultural work visa. The presence of female residents was a dichotomous measure, as was whether a NCDOL inspection certificate had been posted. All camps inspected by the NCDOL should post the inspection certificate; because all camps included in this study housed migrant farmworkers, all should have been inspected.

No definition of how to characterize the visibility of a farmworker labor camp exists in the literature. We determined the hidden status of farmworker labor camps using 2 characteristics. A camp was considered hidden if (1) it was at least 0.15 miles (792 ft or 241.4 m) from a publicly maintained all-weather road or (2) the view of the camp from the road was blocked by a natural or manufactured structure. We used ArcInfo 9.3 software (Esri, Redlands, CA) to quantify the camp distance from a paved road. The distance was calculated using “near analysis” at the Winston-Salem State University Center for Community Safety GIS laboratory. This function is a point (camp) to polyline (road) distance calculation between data layers in the GIS software. The polyline shape file used to represent the road layer in the analysis was published by the North Carolina Department of Transportation.³⁰ We used Google Earth Pro software satellite imagery to determine obstacles that obstructed the camp from public view. The frame of reference for public view was the nearest paved road. An image of the camp that included the road perspective was taken and then coded for obstruction in ATLAS.ti (Scientific Software Development GmbH, Berlin, Germany), a qualitative analysis software application. We used QGIS to make maps for this analysis.

We field-validated the hidden measure by a return visit to a random selection of 10 camps that had been classified as hidden. We also visited another 13 camps en route to the randomly selected 10. All 23 camps were validated by taking pictures of the camps from the road.

Analysis

Descriptive statistics included means and standard deviations for continuous measures and counts and percentages for categorical measures. All statistical tests were 2-sided; significance was determined at the 0.05 probability level. All analyses were performed using SAS version 9.2 (SAS Institute, Cary, NC).

RESULTS

Hidden and visible farmworker labor camps were distributed across the study area (Figure 1). Figures 2 through 4 present examples of camps that included the Google Earth view and an image taken from the road. Figure 2 shows an example of a camp composed of barracks hidden by obstruction. The farmworker labor camp is behind a row of barns used in curing tobacco. Figure 3 shows an example of a camp composed of barracks hidden by distance from the road. The satellite image shows that this camp is not only a great distance from the road but also that it is separated from the road by a wooded area. Figure 4 shows an example of a visible camp. This camp consists of an old farmhouse located near the road.

FIGURE 1—

Location of farmworker labor camps that participated in the study, indicating hidden and visible camps.

FIGURE 2—

Example of farmworker labor camp hidden by obstruction by barns used to cure tobacco, shown from (a) satellite image and (b) nearest public road.

FIGURE 3—

Example of farmworker labor camp hidden by distance, shown from (a) satellite image and (b) nearest public road.

FIGURE 4—

Example of visible farmworker labor camp, shown from (a) satellite image and (b) nearest public road.

Of the 180 camps, 68 (37.8%) were hidden (Table 1). Of the 68 hidden camps, 16 (23.5%) were hidden by distance, 17 (25.0%) were hidden by obstruction, and 35 (51.5%) were hidden by both distance and obstruction. Hidden camps had a mean distance from the road of 0.24 ±0.12 SD miles; visible camps had a mean distance from the road of 0.05 ±0.04 SD miles. Camps hidden by distance alone had a mean distance to the road of 0.26 ±0.07 SD miles; camps hidden by obstruction alone had a mean distance to the road of 0.10 ±0.02 SD miles.

TABLE 1—

Camp Characteristics: East Central North Carolina, 2010

Camp Characteristics	Total, No. (%)	Hidden, No. (%)	Visible, No. (%)	P
All camps	180 (100.0)	68 (37.8)	112 (62.2)
No. of residents				≤ .001
1–6	40 (22.2)	7 (10.3)	33 (29.5)
7–10	44 (24.4)	11 (16.2)	33 (29.5)
11–20	48 (26.7)	21 (30.9)	27 (24.1)
≥ 21	48 (26.7)	29 (42.7)	19 (17.0)
No. of dwellings				.002
1	108 (60.3)	30 (44.1)	78 (70.3)
2	36 (20.1)	18 (26.5)	18 (16.2)
≥ 3	35 (19.6)	20 (29.4)	15 (13.5)
Housing type				≤ .001
Nonbarracks	124 (68.9)	34 (50.0)	90 (80.4)
Barracks	56 (31.1)	34 (50.0)	22 (19.6)
H-2A status				.202
H-2A	124 (68.9)	43 (63.2)	81 (72.3)
Non–H-2A	56 (31.1)	25 (36.8)	31 (27.7)
Female residents present in the camp	46 (25.6)	20 (29.4)	26 (23.2)	.355
NCDOL certificate of inspection				.398
Posted	61 (35.1)	25 (39.1)	36 (32.7)
Not posted	113 (64.9)	39 (60.9)	74 (67.3)

Note. H-2A = temporary agricultural work visa; NCDOL = North Carolina Department of Labor. The sample size was n = 180.

A significantly greater percentage of hidden camps were larger, with more residents (42.7% vs 17.0% with ≥ 21 residents) and buildings (29.4% vs 13.5% with ≥ 3 dwellings). A significantly greater percentage of hidden camps had barracks (50.0% vs 19.6% for visible camps). H-2A visa status, presence of female residents, and the presence of NCDOL inspection certificates did not vary significantly between hidden and visible camps.

DISCUSSION

Poor housing conditions in farmworker labor camps often go unnoticed because their rural location keeps the general population ignorant of their circumstances. Our results indicate that many farmworker labor camps are hidden, thus restricting public knowledge of their conditions. Camps with barracks housing, which are built and designed to house migrant farmworkers, are more likely to be hidden. Building barracks to house farmworkers in a labor camp implies the forethought of where to situate the buildings, pointing to an intentional use of space to hide this vulnerable population. Locating these barracks out of public view hides the existence of the farmworkers. Hidden labor camps do not benefit from community accountability, which could promote the enforcement of housing regulations. Their hidden existence does not enable natural surveillance. Hidden camps increase farmworker structural vulnerability as service providers and advocates working to improve health and justice for farmworkers are not able to contact them.

Although camps with barracks were more often hidden, other analyses have indicated that camps with barracks had fewer migrant housing violations than camps with no barracks.²² The fewer violations and greater distance from the road may have been a function of barracks being newly and specifically built to house migrant farmworkers, as opposed to using old farm houses and trailers. At the same time, fewer camps with barracks and that were hidden afforded farmworkers with privacy, where privacy was defined as dividers between toilets and showers.²⁵ One quarter of camps with no barracks had privacy issues, compared with 83.6% of those with barracks; one third of visible camps had privacy issues, compared with 61.2% of hidden camps.²⁵ This difference may indicate that the lack of privacy is built into hidden camps, as privacy screens are not part of migrant housing regulations and are at the discretion of the employer.

Research on migrant farmworker housing is limited. Research in the peer-reviewed literature has been based on a few studies in scattered locations across the United States, including California,^19,31,32 Oregon,³³ Minnesota,¹⁶ North Carolina,^{22,25,34–37} and Florida.³⁸ National data are available from a survey conducted by the Housing Assistance Council,^39,40 but the results were not peer-reviewed, and the data, collected from 1997 to 2000, may no longer be applicable. The Current Population Survey⁴¹ and 2011–2012 National Agricultural Workers Survey⁴² provide some national data on farmworker housing. A recent analysis indicated that farmworker housing in California is moving from rural to urban locations.⁴³ However, none of the literature we could locate provided a geographical analysis of the location of farmworker housing.

The current spatial norms of farm labor camps hide a large portion of this population, exacerbating the structural vulnerability of farmworkers. Living in hidden housing undermines the social position of farmworkers and is another reminder of their marginalization. Building codes and zoning laws should promote community accountability for farmworker housing using principles of crime prevention through environmental design.²⁶ Because farmworkers are vulnerable, laws must be written that protect their rights.

This analysis should be assessed in light of its limitations. It was based on cross-sectional data collected in a single year in 1 region of North Carolina; it may not be generalizable to other regions, states, or years. We did not contact and interview the growers who own farmworker labor camps about their motives in locating these camps. Our definition of what constitutes a hidden labor camp is temporal in nature; buildings get demolished and roads are developed, which could change the hidden status of a camp. Distance from the main road may not be the best vantage point for community accountability. No research exists as to how an appropriate sight distance affords accountability. Finally, we may not have located some farm labor camps, but these would probably be those that are the most hidden. Their inclusion might add further emphasis to the results we reported.

The analysis had several strengths. It used a novel application of GIS for an evidence-based measure of the spatial status of farmworker labor camps. GIS was used to calculate the distance from the road, and satellite imagery was evaluated to determine the presence of obstructions to visibility of the labor camps. Labor camps included in the analysis were “ground-true” locations. The measure was field-validated with a return trip to a subsample of camps. This research was an example of how CBPR generates a hypothesis of interest to the community. Farmworker outreach workers and community partners are keen to share insights into how they find labor camps that are hidden from public view. For example, they told us that they follow old school buses filled with migrant farmworkers leaving discount store parking lots.

Regulations should be enacted that empower this disenfranchised population, with rules that discourage interaction with farmworkers being repealed. Similar to other instances of environmental injustice, poor housing conditions increase the health risks of migrant farmworkers. Camps that are hidden from public surveillance increase the probability that this vulnerable population will be forced to live in substandard housing that does not meet the minimum standards of current regulations and that farmworkers will be subject to other forms of crime.²³ New North Carolina legislation that deters community engagement in labor camps by classifying trespassing on agricultural facilities as a Class A1 misdemeanor (G.S. 14–159.12(c)(1)) is an example of a regressive policy. Proponents of this policy seek to limit access to farms. Policy should be considered that would promote greater community engagement with farmworker labor camp residents and reduce their structural vulnerability. Laws that work to restore dignity to this essential workforce should include provisions for improved farmworker housing visibility.