Printer Friendly

Undocumented immigration and the business of farm labor contracting in the USA.

Abstract

Purpose--Farm labor contractors operate as intermediaries between farmworkers and agricultural employers by recruiting and supplying labor to US farms. In a political economy where there are employer sanctions for hiring workers without proper documentation, contractors share risk alongside final employers. Furthermore, contractors may facilitate quick employment matches during time sensitive agricultural tasks such as harvesting. For undocumented workers, using a contractor may decrease uncertainty associated with a foreign labor market and ease language barriers. The purpose of this paper is to examine the current role of labor contractors in delivering immigrant agricultural workers, particularly undocumented workers, to farms.

Design/methodology/approach--Determinants of labor contractor use and relationships to final worker outcomes are examined using econometric methods and a large nationally-representative worker survey that is distinctive in that it distinguishes legal status.

Findings--Undocumented farmworkers are shown to be more likely to use contractors than are documented workers, though statistical significance is sensitive to the inclusion of crop and task indicators, and wages and fringe compensation to workers who use contractors are lower, even after controlling for legal stares.

Research limitations/implications--The paper contributes to limited recent academic work on the role of labor contractors in US agriculture. Future work may examine ongoing changes to this role in the context of mumble immigration policy and public opinion.

Practical implications--It is argued that the decline in labor contracting increases the need for employer-level bilingual communication skills and compliance with labor regulations.

Originality/value--Understanding current dynamics of the agricultural labor market should be of value to scholars of rural economies, farm owners and agricultural policymakers.

Keywords United States of America, Farms, Labour utilization, Migrant workers, Farm labor contractors, Migrant and seasonal farmworkers, Illegal immigration, Agricultural management

Paper type Research paper

Introduction

Farm labor contractors operate as intermediaries between farmworkers and agricultural employers by recruiting and supplying agricultural labor to US farms. The business of farm labor contracting can be compared to employment agencies supplying temporary workers in other industries. Some important differences exist, however, Particularly in terms of distinct interrelationships of this type of labor management practice with the undocumented immigrant population in US agriculture.

In a political economy where there are employer sanctions to hiring workers without proper documentation, farm labor contractors, being ultimate suppliers of labor to US farms, can share enforcement risk alongside final employers [1]. This is in addition to any simplifications of the farm operation in terms of worker recruitment and management for often time sensitive tasks associated with harvest cycles. From the worker perspective, using a labor contractor may facilitate a tricky matching process decreasing uncertainty associated with navigating a foreign labor market especially when one is undocumented and even more so in cases where the potential worker is not confident in English language. Value added provided by the farm labor contractor therefore can be seen as twofold benefitting both the growers demanding a workforce quickly and at minimal risk, and the workers looking for fast employment matches regardless of language and labor market familiarity disadvantages.

Previous literature has noted these and other relationships between farm labor contracting and undocumented immigrant populations in US agriculture historically. Polopolus and Emerson (1999), for example, contend that farm labor contractors have advantages over growers in terms of undocumented worker recruitment because of their bilingual skills and understandings of migrant labor supply networks especially undocumented networks. Furthermore, Thilmany (1995) describes informal business relationships between growers and farm labor contractors that are often characterized by a lack of formal contracts. This limited paper trail also is suggestive of a risk minimization in the context of increasing or uncertain immigration enforcement. For the worker perspective, Billikopf (1997) documents perceptions of advantages and disadvantages associated with working for a farm labor contractor in comparison to being directly hired by a grower. Advantages include longer work seasons and assistance with contracts when language is an issue, and disadvantages include less work, lower pay, fewer benefits, and poor work conditions especially when workers are undocumented.

Due to associations with the undocumented population, public policy aiming to decrease undocumented worker inflows in agriculture specifically has and continues to target farm labor contractors. This article therefore examines the ongoing role of farm labor contractors in the business of delivering immigrant agricultural workers, particularly undocumented workers, to US farms using a large nationally-representative worker survey that is distinctive in that it distinguishes legal status. Determinants of farm labor contractor usage as well as relationships to final worker outcomes are examined. Undocumented farmworkers are shown to be more likely to use farm labor contractors than are documented workers though the statistical significance of this result is sensitive to the inclusion of specific crop and task indicators. As may be expected given value added of contractor services, wages paid to workers who used labor contractors are lower even after controlling for legal status as are probabilities of receiving employer-provided housing, bonus payments, and health insurance as supplemental compensation. Finally, discussion focuses on how the prevalence of farm labor contracting in the US has decreased over time while the prevalence of undocumented workers in agriculture has continued to increase. A hypothesis explaining this pattern in light of the overall cross-sectional patterns described is that these changing patterns have occurred simultaneously with strengthening of immigrant personal networks that facilitate immigration and employment on US farms in the absence of intermediaries. The article, in its examination of the business of farm labor contracting, thus contributes to literatures on relationships between immigration and workforce recruitment and management, and on undocumented immigration and agricultural labor in the USA.

Literature review: farm labor contracting in US agriculture

A series of legislation historically in the US has aimed to decrease the practice of recruiting and hiring undocumented workers for US farmwork while protecting agricultural workers from exploitation. Early papers examined the effectiveness toward achieving these goals of the 1963 Farm Labor Contractors Registration Act, which required contractors to file for registration certificates and hold liability insurance for transporting workers, and the 1982 Migrant and Seasonal Workers' Protection Act, which increased recordkeeping and disclosure of employment terms and conditions. These papers, however, found limited evidence of the satisfaction of these objectives (Vaupel and Martin, 1986, 1987). Vaupel and Martin (1987), for example, document that enforcement and employer sanctions have had limited effects on the numbers of undocumented immigrants hired by farm labor contractors despite both pieces of legislation specifically prohibiting the hiring of undocumented workers.

Taylor and Thilmany (1993), in a study of the California farm labor market, find that the 1986 Immigration Reform and Control Act (IRCA), which provided a path to legalization (amnesty) to undocumented immigrants present in the US at the time and increased employer enforcement pertaining to the practice of hiring undocumented workers, also had limited influence on undocumented immigrant hiring and on farm labor contractors. The authors argue that IRCA may have actually increased, not decreased, incentives to use contractors. They hypothesize that in response to new employer sanctions, employers are left with the choice of either hiring workers on their own directly and then investing in reducing apprehension risk (for example, via reducing turnover of current employees), or recruiting their workforce via farm labor contractors in order to share risk of employer sanctions with these second employing agents. The authors, however, do not find substantial evidence of reduced turnover in support of the first mechanism but do find some evidence suggestive of a role of the second. Patterns of increased farm labor contractor usage also were predicted by Martin and Taylor (1990) who cited survey evidence indicating employer intent to use contractors to a greater extent following a negative supply shock and described in Martin (1994) as an unintentional effect of immigration reform. Aggregating across occupations, Phillips and Massey (1999) document an increasing wage penalty for those hired via a contractor.

Since more recent academic work on the role of farm labor contractors is limited, a focus of this article is to examine ongoing determinants of farm labor contractor usage and relationships to worker outcomes measured in terms of wages received using repeat cross-sectional data over recent years. A final discussion question is to consider what survey evidence can reveal about the extent to which the role of farm labor contractors in US agriculture has changed since the earlier literature and in the context of continuing changes in immigration policies and public opinion atmosphere.

Data: The National Agricultural Workers Survey

Cross-section data come from the US Department of Labor's National Agricultural Workers Survey (NAWS), a nationally and regionally representative survey of employed US farmworkers with survey weights. Farmwork regions are based on US Department of Agriculture definitions. Workers are randomly selected from work sites when arriving for work, at lunch, or when leaving, and interviews are scheduled for times, at locations, and in languages chosen by the workers. Survey respondents have been sampled in three seasons (approximating agricultural seasons) per year since 1989, and the total sample size through the 2006 wave is 46,566. Questions pertain to general demographic and work-related characteristics, personal and family histories, as well as detailed information on legal status [2].

Farm labor contractor usage in the NAWS

Of the weighted data and across survey years, 19.55 percent of workers overall report being employed via a farm labor contractor as opposed to being directly hired by a grower. Differences exist across legal status groups. For example, only 11.67 percent of naturalized citizens reported that their employer at the time of the interview was a farm labor contractor, compared with 21.39 percent of green card holders and those with other work authorizations. Of undocumented workers, the percentage is even greater; 27.16 percent used a farm labor contractor to achieve their current job. For comparison, only 5.94 of native, US born hired agricultural workers report employment via a labor contractor.

Figure 1 shows the overall prevalence of farm labor contractor usage among immigrant workers over the 1989-2006 period. Highest rates for farm labor contracting are seen for the case of undocumented workers, followed by those with legal immigrant status defined to include naturalized citizenship, green cards, and other forms of work authorization. Although variable across years, this difference between undocumented and documented workers persists across the time series.

Martin and Miller (1993) indicate an increase in the farm labor contracted employment from the 1984 to 1990 period for the California case. While this same trend is evident in the raw national data from the NAWS as shown in Figure 1, similar increases and equal magnitude decreases are shown throughout the series suggesting that as a percentage of the total farm labor force, the prevalence of farm labor contracting across the longer term has tended to revert to approximately 20 percent before decreasing further toward the end of the series. These patterns, however, are unconditional and therefore do not account for other demographic and work-related characteristics.

[FIGURE 1 OMITTED]

Demographic and work-related characteristics of employed farmworkers

Summary statistics of demographic and work-related characteristics of employed farmworkers across two groups based on farm labor contractor usage versus direct hire reveal statistical significant differences. As presented in Table I, those who used a farm labor contractor were less likely to be female, were younger on average and with less education, experience, and US family ties. These workers were most likely to be undocumented among legal status categories identifiable in the data, and were often engaged in fruit and vegetable harvest picking operations and observed in the California agricultural region. These patterns are not surprising given hypotheses that farm labor contractors serve labor recruitment and employment matching purposes. Younger, less educated, inexperienced workers without legal status and with less established personal and family networks, for example, theoretically can benefit most from labor contracting services. Notably and consistent with literature on interplays between farm labor contracting and language skills, only 15 percent of workers using farm labor contractors report high English language speaking skills (and less than 14 percent for reading). This is in comparison to approximately 34 percent for speaking (and 30 percent for reading) for those not using farm labor contractors to secure employment. These differences are highly statistically significant.

Empirical framework and results

Empirical methodology involves examining determinants of farm labor contractor usage with attention to legal status characteristics of workers followed by identifying causal effects of farm labor contractor usage on wages received by workers while controlling for legal status and other demographic and work-related characteristics directly.

Determinants of farm labor contractor usage

The determinants of farm labor contract usage are presented in Table II. Probit marginal effects are reported from a regression in which the probability of having used a farm labor contractor is modeled as a function of demographic and work-related characteristics.

Results are presented in a stepwise fashion with least restrictions presented first. Notably, in the absence of crop and task controls, indicators of legal status are highly statistically significant. For example, undocumented status is associated with approximately 10-20 percent higher propensities to use a farm labor contractor over being directly hired independent of particular agricultural job performed. Once crop and task characteristics are included, however, legal status variables are not significant. This is consistent with the prevalence of employers relying on farm labor contractors for the purpose of achieving particular, often temporary and time sensitive, tasks for specific crop types within the agricultural season, and with correlations between undocumented labor and particular crops. In this context, legal status may be of reduced importance to growers given the time sensitivity of certain agricultural tasks[3].

While the operative mechanism is seen to stem from crop and task attributes, regionality also is of relevance with all regions (coefficients not shown) demonstrating negative and highly statistically significant differences over the base (excluded) category of California in the regression presented in column (4). This can be interpreted as relating to the importance of farm labor contracting in the California region as demonstrated in the summary statistics presented in Table I where more than 56 percent of workers reporting farm labor contractor usage were surveyed in California in comparison to only 26 percent of those not reporting labor contractor use. According to the Economic Research Service of USDA, California accounted for 51 percent of the harvested fruit and tree nut acreage in the US and 25 percent of the harvested area of vegetables and melons in addition to being the top state producer in terms of quantity. Fruit and vegetable harvesting is time sensitive and temporary and therefore especially suitable for sustaining farm labor contracting business.

Effect or farm labor contract usage on individual worker compensation

Farm labor contractors perform a particular service as intermediary and facilitator and therefore wage differentials between those directly hired and hired via the contractor are expected. Specifically, wages for workers relying on labor contractors as opposed to direct hire theoretically would be lower as the result of both costs associated with the matching service provided by the contractor and lower worker bargaining powers at an individual level. This empirical pattern has been documented in case studies back to Rooney (1961). Simple wage regressions confirm this finding in the NAWS data which extends to the more current time period.

The effect of farm labor contractor usage on wages can be modeled in a multivariate framework. This basic econometric framework takes the general form:

Log(wage)i = [alpha][FLC.sub.i] + [X.sub.i][beta]+ [[epsilon].sub.i] (1)

where the vector [X.sub.i] includes legal status, nativity, and general demographic and work-related characteristics such as gender, age, education, experience, tenure, crop, task, geographic region of observation, and survey year. Of particular interest is the statistical and economic significance of the coefficient a on the farm labor contractor usage variable.

Since some agricultural workers are paid piece rates (i.e. wages based on output) instead of time rates (i.e. wages based on time input), hourly-equivalent wages are constructed for these workers based on survey responses indicating how much the worker (and his or her crew if applicable) was paid on average for each unit of output (e.g. box, bin, etc.), how many units were produced in an average day, and crew size (if paid as a group). Ordinary least squares estimates of the effects of farm labor contractor usage on hourly and hourly-equivalent farmworker wages are presented in Table III, and the effect is estimated at -2.5 to -3.6 percent. The difference between farmworker wages from a farm labor contractor and from direct hire by a grower can be hypothesized to accrue to the labor contractor and grower by some split determined by their relative bargaining powers in the labor arrangement at their level.

A related question is to what extent estimated hourly or hourly-equivalent wage differentials reflect full differences in compensation received by farmworkers using and not using farm labor contractors. Hourly and hourly-equivalent wage differentials as estimated can be thought of as lower bounds on full compensation differentials between those using a farm labor contractor versus those directly hired by a grower as differentials accounting for in-kind transfers (such as housing), bonus payments, and insurance operate in the direction of increasing the true compensation gap.

Other data within the NAWS can help shed light on this question. First, workers are asked about their current housing situations. Of the weighted sample, 14.4 percent of those working via a labor contractor report rent-free employer-provided housing. Of those working via a grower directly, 22.5 percent report receiving free housing. Similarly, 6.2 percent of those working for a contractor and 27.4 percent of those working for a grower report receiving a bonus in addition to regular wage compensation. Further differentials exist in terms of additional benefits such as health insurance. Of farmworkers under labor contractors 65.7 percent report that they have access to employer-provided health insurance or payment for health care for injuries and illness originating at the worksite and 3.2 percent report the same for occurrences outside of the workplace. These numbers can be compared with 75.2 and 11.1, respectively, for workers directly hired by growers.

Table IV presents coefficients on farm labor contractor use for probit regressions similar to equation (1) but taking the probability of employer-provided housing, bonus payments, and health insurance (defined to include injuries and illness outside of the worksite), respectively, as the dependent variable. Regressors parallel those in columns (1) through (4) in Table III. Farm labor contractor usage is associated with 5-10 percent lower probabilities of free housing all else equal, 13-16 percent lower probabilities of bonus payments, and 5-5.5 percent lower probabilities of employer-provided health insurance.

Selection bias and propensity score matching

Estimation by OLS may not be appropriate if any selection bias exists. In order to account for potential selection bias and identify causal effects, farm labor contractor usage is modeled here using the semi-parametric technique of propensity score matching. Propensity score matching methods are based on balancing observable characteristics in the data and rely on fewer distributional assumptions than traditional parametric methods (Dehejia and Wahba, 2002). Specifically, consider the propensity to use a contractor as an unobserved latent variable:

[FLC.sup.*.sub.i] = [z.sub.i][gamma] + [u.sub.i] (2)

where the treatment decision rule is:

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Equation (2) can be interpreted as similar to the exercise in Table II that presented estimates of probit marginal effects of various demographic and labor market characteristics on the probability that a worker used a farm labor contractor.

To identify the effects of farm labor contractor usage on wages, workers are matched based on a propensity score or measure of their observed characteristics. The first step of the general technique is to estimate an equation similar to equation (2) in order to generate a predicted value of the dependent variable. The second step is to examine the effect of farm labor contractors on wages by matching treatment and control observations based on the predicted values and creating counterfactuals. Here, propensity scores are based on covariates used in previous OLS estimations. The balancing property is satisfied, meaning that the data are sufficient to successfully form treatment and control groups, for ten blocks based on gender, naturalized citizen status, green card possession, other authorization, undocumented legal status, Mexican and Central American origin, and work in Midwest and Northwest agricultural regions within the USA[4].

To construct counterfactuals, matching is performed based on both individual neighborhood and on smooth weighting in subsequent estimations. For individual neighborhood, or nearest-neighbor, matching is based on grouping observations that are closely ranked together by the propensity score measure. For smooth weighting, here a kernel matching method, matching is based on an assumed population distribution. For both, matching is performed within the common support region meaning that there is overlap between the distribution of observed characteristics in the control and treatment groups.

An identifying assumption is that treatment and control observations with like propensity scores differ only in the error term from the first equation. The main average treatment on the treated estimation is then simply calculated as:

E(Log[(wage).sub.i1] - (Log[(wage).sub.i0]|[FLC.sub.i] = 1) = E(Log[(wage).sub.i1]|[FLC.sub.i] = 1) - E(Log[(wage).sub.i0]|[FLC.sub.i] = 1) (3)

where E(Log[(wage).sub.i1]|[FLC.sub.i] = 1) and E(Log[(wage).sub.i0]|[FLC.sub.i] = 1) are actual and counterfactual average hourly or hourly-equivalent wages for the cases that workers did or did not use a contractor to secure employment. The probabilities of housing, bonuses, and health insurance replace Log(wage) in the calculations for other compensation.

Results follow the general patterns identified by the OLS technique though effects of farm labor contractor usage on wages received are found to be economically larger after controlling for selection. Farmworkers who are employed by a farm labor contractor earn an estimated 4.4 percent lower hourly wage than workers employed directly by a grower all else equal. This result appears for both matching methods presented and therefore is not sensitive to choice of matching method. By the propensity score matching method, farmworkers who are employed by a labor contractor have 8.8-9.1 percent lower probabilities of receiving employer-provided housing, 18-18.8 percent lower probabilities of bonuses, and 7.5-7.7 percent lower probabilities of employer-provided health insurance than workers hired directly by a grower. Thus, by not directly accounting for selection, OLS underestimates the effect of farm labor contracting on wages, bonuses, and health insurance benefits but provides similar estimates for propensities to receive employer-provided housing (Table V).

Discussion and conclusions

Farm labor contractors in US agriculture facilitate job matching between growers and prospective field laborers. Of legal status groups, undocumented immigrants are mostly likely to report using a farm labor contractor. This makes sense if the gains to using a labor contractor are increased for undocumented immigrants who have limited experience with US labor markets, and labor contractors provide a risk sharing service to the growers to which they supply labor. Employment via a farm labor contractor is found to be associated with between two and four percent lower wages than via direct hire overall once individual and job-related characteristics are taken into account and simple OLS is used as an estimation method. When controlling for selection by using propensity score methods, the effect is estimated in the negative four to five percent differential range. The effects of farm labor contractor usage on the probabilities of receiving bonuses and employer-provided health insurance benefits also are underestimated by OLS. Accounting for selection bias, these effects are estimated at -18 to -19 and -7.5 to -8 percent, respectively. The effect of labor contracting on employer-provided housing is estimated at approximately nine percent in the negative direction.

Farm labor contracting over time

Previous literature has indicated that demand for farm labor contracting is a positive function of legal immigration enforcement due to employer level incentives to substitute direct hire practice for outsourced labor contracting services. On the supply side, Vandeman et al. (1999) predict that decreases in the flow of undocumented workers would lower propensities to use farm labor contractors and increase wages. Despite policy changes in the direction of increased enforcement continuing throughout the period corresponding to the nationally representative data used in this article, the fraction of the farmwork population that is undocumented has increased over time while the fraction reporting usage of farm labor contractors has decreased. Levine (2004) documents similar patterns of increasing proportions of workers directly hired by growers using Current Population Survey and the US Department of Agriculture's Farm Labor Survey data. Figure 2 shows the fraction of farmworkers overall (immigrant and native combined) who report using farm labor contractors alongside the fraction of the agricultural worker population that is undocumented over survey years in the NAWS sample. The inverse pattern is notable given previous literature on interrelationships and the cross-sectional description here. Earlier work and federal policy has taken the opinion that farm labor contracting as a business fuels undocumented immigration and not the other way around. Recent trends, however, suggest that undocumented immigration is sustained at a high level even as labor contracting diminishes.

[FIGURE 2 OMITTED]

Although not directly testable here, both supply- and demand-side hypotheses explaining the pattern shown in Figure 2 are reasonable. If enforcement changes over time increase expected business costs of farm labor contractors, for example, then the response may be on the supply side of the labor contracting market. As noted, however, previous studies testing this hypothesis following the 1986 Immigration Reform and Control Act, which increased enforcement, failed to find supportive evidence of this dynamic. While unfortunately the NAWS, given its year coverage, does not afford the opportunity to study the effects of IRCA, other pertinent immigration legislation should be mentioned. The Illegal Immigration Reform and Immigrant Responsibility Act of 1996 (IIRIRA) which went into effect in mid-1997 and increased general enforcement especially that pertaining to deportation corresponds to the approximate time that the farm labor contracting series begins to show an alternate picture in Figure 2. Furthermore, the Migrant and Seasonal Agricultural Worker Protection Act was restated concurrent with IIRIRA to more clearly define joint employment between a grower and a farm labor contractor for legal purposes facilitating regulation enforcement.

While these legislation introductions and clarification provide a supply-side hypothesis consistent with the patterns in Figure 2, alternately it is possible that the illustrated changing patterns have occurred simultaneously with strengthening and further development of immigrant personal networks. If undocumented immigrants have access to networks that facilitate finding US employment (either within or outside of agriculture) at lower costs than those associated with farm labor contractors then the advantage of using a farm labor contractor is diminished. In this case, the response is a demand side one. Examining precise aspects of structural change in agricultural labor markets is left for future work, as is examining ongoing changes to this role of contractors in the context of mutable immigration policy and public opinion.

Managerial and policy implications

In a case study of California citrus, Mines and Martin (1984) described unionized workers as being less attractive to farm owners than contracted workers because of costs involved, and Polopolus and Emerson (1999) hypothesized that the use of farm labor contractors was related to business organizational practice changes with the goal of minimizing the threat of employer sanctions. Being consistent with a long-term profit motive, farm labor contracting therefore should increase with lack of enforcement both in terms of labor regulations and in terms of immigration policy and likewise decrease in the presence of enforcement.

According to the 2007 Census of Agriculture (United States Department of Agriculture, 2009), 182,701 US farms used contract labor. This number is down from 228,692 by the 2002 Census. Total expenses on contract labor, however, were up, totaling $3.45 billion in 2002 and $4.51 billion in 2007 (2.0 and 1.9 percent of total farm expenses, respectively). Figure 3 shows contract labor expenses by farm size, and Figure 4 shows total numbers of farms employing contract labor by acreage. As expected, expenses on contract labor in Figure 3 are correlated with farm size with higher expenses accruing to the larger farms likely due to magnitudes of farm task operations. Figure 4 adds to the story indicating that 48 percent of farms employing contract labor had less than 500 acres compared with 20 percent with 500-999 acres, 16 percent with 1,000-1,999 acres, and 19 percent with acreage above 2000. This suggests that smaller farms may actually be more dependent on labor contractors than are larger farms. A decrease in farm labor contracting over time therefore has implications for small farm ownership.

[FIGURE 3 OMITTED]

[FIGURE 4 OMITTED]

Facing a smaller farm labor contractor market, small farm owners may find that substitutions toward investments in their own labor management and leadership are attractive. A relative absence of labor contractors, for example, increases the need for bilingual communication skills to continue hiring laborers matching today's demographic and increases the need for familiarity and compliance with labor regulations if growers are risk averse and previously used labor contractors to hedge enforcement risk. Alternatively (or in addition), farm owners can make greater use of temporary work visa programs especially the underused federal H-2A program in conjunction with state level initiatives offering labor recruitment mechanisms, or further mechanize their harvest operations via greater reliance on new technologies.

References

Billikopf, G.E. (1997), "Workers prefer growers over FLCs", California Agriculture, Vol. 51 No. 1, pp. 30-2.

Dehejia, R. and Wahba, S. (2002), "Propensity score-matching methods for nonexperimental causal studies", The Review of Economics and Statistics, Vol. 84, pp. 151-61.

Levine, L. (2004), Farm Labor Shortages and Immigration Policy, Paper 23, Federal Publications, available at: http://digitalcommons.ilr.cornell.edu/key_workplace/23

Martin, P.L. (1994), "Good intentions gone awry: IRCA and US agriculture", Annals of the American Academy of Political and Social Science, Vol. 534, pp. 44-57.

Martin, P.L. and Miller, G.P. (1993), <Farmers increase hiring through labor contractors", California Agriculture, Vol. 47 No. 4, pp. 20-3.

Martin, P.L. and Taylor, J.E. (1990), "The initial effects of immigration reform on farm labor in California", Population Research and Policy Review, Vol. 9 No. 3, pp. 255-83.

Mines, R. and Martin, P.L. (1984), "Immigrant workers and the California citrus industry", Industrial Relations, Vol. 23 No. 1, pp. 139-49.

Phillips, J.A. and Massey, D.S. (1999), "The new labor market: immigrants and wages after IRCA", Demography, Vol. 36 No. 2, pp. 233-46.

Polopolus, L.C. and Emerson, R.D. (1999), "Entrepreneurship, sanctions, and labor contracting", Southern Journal of Agricultural Economics, Vol. 23 No. 1, pp. 57-68.

Rooney, J.F. (1961), "The effects of imported Mexican farm labor in a California county", American Journal of Economics and Sociology, Vol. 20 No. 5, pp. 513-21.

Taylor, J.E. and Thilmany, D. (1993), "Worker turnover, farm labor contractors, and IRCA's impact on the California farm labor market", American Journal of Agricultural Economics, Vol. 75 No. 2, pp. 350-60.

Thilmany, D. (1995), "An analysis of contract relationships between farm labor contractors and farmers in California agriculture", Publication APMP004, Agricultural Personnel Management Program, University of California Cooperative Extension, Division of Agriculture and Natural Resources.

United States Department of Agriculture (2009), United States Summary and State Data, Volume 1, Geographic Area Series, Part 51, AC-07-A-51.

Vandeman, A., Sadoulet, E. and de Janvry, A. (1999), "Labor contracting and a theory of contract choice in California agriculture", American Journal of Agricultural Economics, Vol. 73 No. 3, pp. 681-92.

Vaupel, S. and Martin, P.L. (1986), "Farm labor contractors", California Agriculture, Vol. 40 No. 3, pp. 12-15.

Vaupel, S. and Martin, P.L. (1987), "Evaluating employer sanctions: farm labor contractor experience", Industrial Relations, Vol. 26 No. 3, pp. 304-13.

Wells, M.J. and Villarejo, D. (2004), "State structures and social movement strategies: the shaping of farm labor protections in California", Politics & Society, Vol. 32, pp. 291-326.

Anita Alves Pena

Department of Economics, Colorado State University, Fort Collins, Colorado, USA

Notes

[1.] An exception is that some state-level labor regulations hold only growers and farm operators responsible as the employers of agricultural labor. This is in contrast to federal level policy. For example, see discussion of California's Agricultural Labor Relations Act in Wells and Villarejo (2004).

[2.] Approximately 1 percent (416 workers) declined to answer legal status questions. Respondents receive a pledge of confidentiality and a nominal financial incentive for participation. The NAWS has a long, visible history within farming communities, and the survey design incorporates data validation exercises and adjustments if necessary pertaining to the legal status questions.

[3.] It should be noted that this is suggestive of reverse causality or simultaneity bias if, for example, the role of the farm labor contractor is to sort workers into particular crop and job types. Models presented in the following section controls for selection that may be associated with this type of mechanism.

[4.] A disadvantage of balancing only being achievable based on a subset of observable characteristics is that it is possible that remaining error in the propensity score equation is correlated with the error of interest, and therefore selection bias, though reduced, continues to be present.

Dr Anita Alves Pena is an Assistant Professor of Economics at Colorado State University. Her research interests are in public sector economics, labor economics, and economic development and her current research relates to undocumented and documented immigration, public policy, poverty, and agricultural labor markets. She received her PhD in Economics from Stanford University in 2007, MA in Economics from Stanford University in 2004, and BA in Economics from the Johns Hopkins University in 2001. She teaches in the areas of microeconomics, econometrics, and public finance. Anita Alves Pena can be contacted at: anita.pena@colostate.edu

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com Or visit our web site for further details: www.emeraldinsight.com/reprints

DOI 10.1108/19355181211217616
Table I.
Characteristics of
farmworkers, by farm
labor contractor usage

                            Used FLC   Did not use FLC   Difference

Female (%)                     18.01             22.02         ***
Age (years)                    30.63             32.50
Education (years)               6.27              7.26
Farm experience (years)         7.59              9.56
Tenure (years)                  2.51              4.21
Has spouse (%)                 27.77             37.17         ***
Children (number)               0.63              0.78
US-born (%)                     6.87             22.62
Naturalized citizen (%)         2.33              4.47         ***
Green card (%)                 23.79             23.35
Other authorization (%)         7.10              6.13         ***
Illegal (%)                    59.92             43.43
Speaks English (%)             15.05             33.61
Reads English (%)              13.69             30.44         ***
Mexican (%)                    88.07             70.58
Central American (%)            2.67              3.14
Puerto Rico (%)                 0.30              2.12         ***
Field crops (%)                13.17             16.49
Fruit (%)                      46.16             30.46         ***
Horticulture (%)                1.33             18.34
Vegetables (%)                 37.15             27.29         ***
Misc. (%)                       2.17              7.31         ***
Pre-harvest (%)                20.32             17.99
Harvest (%)                    46.58             31.51         ***
Post-harvest (%)                6.04             12.43         ***
Semi-skill (%)                 21.67             20.11
Supervisor (%)                  0.11              0.28
Other task (%)                  5.29             17.68         ***
California (%)                 56.11             26.10         ***
East (%)                        8.39             18.55         ***
Southeast (%)                  18.32             12.57         ***
Midwest (%)                     5.03             21.33
Southwest (%)                   8.17              7.73
Northwest (%)                   3.99             13.73         ***
Observations                7,920            33,675

Note: Significant at: * p < 0.1, ** p < 0.05 and *** p < 0.01

Table II.
Determinants of farm
labor contractor usage

                             (1)              (2)

Female                  -0.00524         -0.00513
                         (.00765)        -0.0085
Age                     -0.00107 ***     -0.00115 ***
                        (0.000340)       (0.000382)
Education               -0.00539 ***     -0.00457 ***
                        (0.00102)        (0.00119)
Farm experience          0.00204 ***      0.00155 ***
                        (0.00467)        (0.000524)
Tenure                  -0.0144 ***      -0.0156 ***
                        (0.000957)       (0.00109)
Naturalized citizen      0.112 ***        0.0372
                        (0.0210)         (0.0259)
Green card               0.208 ***        0.106 **
                        (0.0147)         (0.0342)
Other authorization      0.198 ***        0.0937 **
                        (0.0205)         (0.0391)
Illegal                  0.204 ***        0.101 ***
                        (0.0116)         (0.0319)
Speaks English                           -0.0682 ***
                                         (0.0142)
Reads English                             0.0200
                                         (0.0182)
From Mexico                               0.0341
                                         (0.0263)
From Central America                     -0.0448
                                         (0.0275)
Field crops

Fruit

Horticulture

Vegetables

Pre-harvest

Harvest

Post-harvest

Semi-skill

Region controls?        No               No
Survey year controls?   No               No
Observations            44,631           41,673

                             (3)                     (4)

Female                      0.00271                 -0.00359
                           (0.00828)                (0.00736)
Age                         4.29 x [10.sup.-5]       0.000286
                           (0.000365)               (0.000342)
Education                  -0.00124                 -0.000435
                           (0.00111)                (0.000987)
Farm experience            -0.000256                -0.000551
                           (0.000492)               (0.000456)
Tenure                     -0.0132 ***              -0.0122 ***
                           (0.00100)                (0.000945)
Naturalized citizen        -0.0101                  -0.0168
                           (0.0215)                 (0.0200)
Green card                  0.0382                   0.00621
                           (0.0288)                 (0.0269)
Other authorization         0.0252                  -0.00123
                           (0.0316)                 (0.0293)
Illegal                     0.0422                   0.0401
                           (0.0287)                 (0.0286)
Speaks English             -0.0535 ***              -0.0215 *
                           (0.0137)                 (0.0129)
Reads English               0.0267                   0.0207
                           (0.0176)                 (0.0158)
From Mexico                 0.0498 **                0.0357
                           (0.0224)                 (0.0235)
From Central America        0.0291                   0.00526
                           (0.0320)                 -0.030
Field crops                 0.137 ***                0.138 ***
                           (0.0202)                 (0.0211)
Fruit                       0.201 ***                0.105 ***
                           (0.0172)                 (0.0166)
Horticulture               -0.123 ***                0.115 ***
                           (0.0129)                 (0.0119)
Vegetables                  0.206 ***                0.143 ***
                           (0.0186)                 (0.0179)
Pre-harvest                 0.105 ***                0.0792 ***
                           (0.0151)                 (0.0145)
Harvest                     0.0755 ***               0.0773 ***
                           (0.0128)                 (0.0126)
Post-harvest               -0.0349 **               -0.0179
                           (0.0137)                 (0.0140)
Semi-skill                  0.0737 ***               0.0556 ***
                           (0.0142)                 (0.0137)
Region controls?         No                       Yes
Survey year controls?    No                       Yes
Observations             41,633                   41,633

Notes: Significant at: * p < 0.1, ** p < 0.05 and *** p < 0.01; robust
standard errors in parentheses; probit marginal effects

Table III.
Effect of farm labor
contractor usage on
log(wages), OLS

                                  (1)                (2)

Used FLC                       -0.0364 ***         -0.0342 ***
                               (0.00652)           (0.00649)
Female                         -0.0525 ***         -0.0547 ***
                               (0.00661)           (0.00657)
Age                             0.000990 ***        0.00974 ***
                               (0.000302)          -0.00135
Age squared/1,000                                  -0.119 ***
                                                   (0.0169)
Education                       0.0127 ***          0.0126 ***
                               (0.00100)           (0.000996)
Farm experience                 0.00235 ***         0.00910 ***
                               (0.000427)          (0.000987)
Experience squared/1,000                           -0.172 ***
                                                   (0.0223)
Tenure                          0.0118 ***          0.0112 ***
                               (0.000637)          (0.000624)
Naturalized citizen            -0.120 ***          -0.129 ***
                               (0.0159)            (0.0157)
Green card                     -0.131 ***          -0.145 ***
                               (0.0177)            (0.0175)
Other authorization            -0.265 ***          -0.271 ***
                               (0.0206)            (0.0204)
Illegal                        -0.140 ***          -0.124 ***
                               (0.0195)            (0.0194)
Speaks English                  0.0165              0.0167
                               (0.0151)            (0.0149)
Reads English                  -0.0145             -0.0158
                               (0.0152) *          (0.0150)
From Mexico                     0.194 *             0.177 ***
                               (0.0157)            (0.0156)
From Central America            0.193 ***           0.179 ***
                               (0.0196)            (0.0196)
Field crops

Fruit

Horticulture

Vegetables

Pre-harvest

Harvest

Post-harvest

Semi-skill

Constant                        1.609 ***           1.450 ***
                               (0.0168)            (0.0254)
Region controls?           No                  No
Survey year controls?      No                  No
Observations               40,446              40,446
[R.sup.2]                       0.108               0.123

                                     (3)                   (4)

Used FLC                       -0.0261 ***               -0.0245
                               (0.00590)                 (0.00589)
Female                         -0.0564 ***               -0.0578 ***
                               (0.00580)                 (0.00580)
Age                            -1.48 x [10.sup.-5]        0.00703
                               (0.000261)                (0.00119)
Age squared/1,000                                        -0.0956 ***
                                                         (0.0148)
Education                       0.00735 ***               0.008
                               (0.000883)                (0.000877)
Farm experience                 0.00185 ***               0.00696 ***
                               (0.000373)                (0.000857)
Experience squared/1,000                                 -0.129 ***

Tenure                          0.00829 ***               0.00783 ***
                               (0.000523)                (0.000517)
Naturalized citizen            -0.0402 ***               -0.0467 ***
                               (0.0125)                  (0.0123)
Green card                     -0.0565 ***               -0.0670 ***
                               (0.0147)                  (0.0145)
Other authorization            -0.0616 ***               -0.068 ***
                               (0.0191)                  (0.0190)
Illegal                        -0.109 ***                -0.0966 ***
                               (0.0161)                  (0.0159)
Speaks English                  0.0315 **                 0.0316
                               (0.0132)                  (0.0130)
Reads English                   0.000916                 -0.000217
                               (0.0129)                  (0.0127)
From Mexico                     0.0991 ***                0.0869 ***
                               (0.0130)                  (0.0128)
From Central America            0.127 ***                 0.117 ***
                               (0.0157)                  (0.0156)
Field crops                    -0.0327                   -0.0302 ***
                               (0.00922)                 (0.00909)
Fruit                          -0.0250 ***               -0.0246 ***
                               (0.00933)                 (0.00922)
Horticulture                    0.00456                   0.00509
                               (0.00810)                 (0.00795)
Vegetables                     -0.0302 ***               -0.0295 ***
                               (0.00850)                 (0.00841)
Pre-harvest                    -0.0602 ***               -0.0588 ***
                               (0.00613)                 (0.00597)
Harvest                         0.0504 ***                0.0505 ***
                               (0.00688)                 (0.00677)
Post-harvest                   -0.00942                  -0.0108
                               (0.00812)                 (0.00804)
Semi-skill                     -0.0284 ***               -0.0305 ***
                               (0.00699)                 (0.00690)
Constant                        1.555                     1.427
                               (0.0245)                  (0.0292)
Region controls?           Yes                       Yes
Survey year controls?      Yes                       Yes
Observations               40,413                    40,413
[R.sup.2]                       0.330                     0.339

Notes: Significant at: * p < 0.1, ** p < 0.05 and *** p < 0.01; robust
standard errors in parentheses

Table IV.
Effect of farm labor
contractor usage on the
probability of receiving
other compensation, OLS

                       (1)           (2)           (3)

Housing            -0.104 ***    -0.104 ***    -0.0493 ***
                   (0.00750)     (0.00748)     (0.00707)
Bonuses            -0.159 ***    -0.157 ***    -0.136 ***
                   (0.00605)     (0.00609)     (0.00642)
Health insurance   -0.0545 ***   -0.0535 ***   -0.0504 ***
                   (0.00445)     (0.00444)     (0.00440)

                       (4)

Housing            -0.0492 ***
                   (0.00707)
Bonuses            -0.133 ***
                   (0.00645)
Health insurance   -0.0495 ***
                   (0.00439)

Notes: Significant at: * p < 0.1, ** p < 0.05 and *** p < 0.01; robust
standard errors in parentheses

Table V.
Effect of farm labor
contractor usage on
worker outcomes,
propensity score
matching

                         Nearest-neighbor      Kernel match

Log(wages)                   -0.044 ***         -0.044 ***
                             (0.004)            (0.004)
Treatment observations    7,935              7,935
Control observations     32,754              33688
Housing                      -0.088 ***         -0.091 ***
                             (0.004)            (0.004)
Treatment observations    7,935              7,935
Control observations     33,553             33,688
Bonuses                      -0.180 ***         -0.188 ***
                             (0.004)            (0.004)
Treatment observations    7,935              7,935
Control observations     31,898             33,688
Health insurance             -0.075 ***         -0.077 ***
                             (0.004)            (0.004)
Treatment observations    7,935              7,935
Control observations     25,662             33,688

Notes: Significant at: * p < 0.1, ** p < 0.05 and *** p < 0.01;
analytical standard errors are reported in parentheses for nearest-
neighbor match method; bootstrapped standard errors are reported for
kernel match method
COPYRIGHT 2012 Emerald Group Publishing, Ltd.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2012 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Pena, Anita Alves
Publication:American Journal of Business
Article Type:Report
Geographic Code:1USA
Date:Mar 22, 2012
Words:6982
Previous Article:Immigration Policy 2.0.
Next Article:Latino informal immigrant entrepreneurs in South Texas: opportunities and challenges for unauthorized new venture creation and persistence.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters