Advertisement
ASHS 2024 Annual Conference

 

What Firms Hire H-2A Workers? Evidence from the US Ornamental Horticulture Industry

Authors:
Xuan Wei Food and Resource Economics Department and Mid-Florida Research and Education Center, University of Florida, Apopka, FL 32703, USA

Search for other papers by Xuan Wei in
This Site
Google Scholar
Close
,
Benjamin L. Campbell Department of Agricultural and Applied Economics, University of Georgia, Athens, GA 30602, USA

Search for other papers by Benjamin L. Campbell in
This Site
Google Scholar
Close
,
Hayk Khachatryan Food and Resource Economics Department and Mid-Florida Research and Education Center, University of Florida, Apopka, FL 32703, USA

Search for other papers by Hayk Khachatryan in
This Site
Google Scholar
Close
, and
Robin G. Brumfield Department of Agricultural, Food & Resource Economics, Rutgers, The State University of New Jersey, New Brunswick, NJ 08901, USA

Search for other papers by Robin G. Brumfield in
This Site
Google Scholar
Close

Abstract

The ornamental horticulture industry relies on workers to do myriad tasks, such as pruning, applying fertilizers, scouting, spraying pesticides, planting, harvesting, packing, and weeding. As a result of the perishable nature of horticultural goods, a skilled and accessible labor supply is imperative for continued industry growth and stability. The decreasing number of workers, followed by increasing wage rates, could be alarming for the economic well-being of the ornamental horticulture industry, which has already experienced a downward trend in revenue and profits. Combining 2014 and 2019 National Green Industry Survey data, this study investigates factors affecting ornamental growers’ decisions on hiring H-2A workers. Growers’ decisions are largely affected by their home state’s enforcement of the 287(g) program and the observed industry employment and total wage payment. Growers are more likely to participate in an H-2A workers program if their home state implemented the restrictive 287(g) program. Increasing industry employment of domestic workers will discourage participation in the H-2A workers program, but increasing industry wage costs will encourage participation. In contrast, individual firm characteristics play different roles in program participation and the number of H-2A workers hiring decision. Increasing farm sales value by $1 million merely increases the probability of hiring by a 0.1% point, revealing that large growers are the major beneficiaries of the H-2A workers program. After the participation hurdle is overcome, the number of H-2A workers hired is affected minimally by these factors. Our results suggest that the current H-2A program imposes a potential hurdle to participate, thus benefiting large growers.

The US agricultural workforce has long consisted of two groups of workers: 1) self-employed farm operators and their family members, and 2) hired workers. Both groups of agricultural workers declined from 1950 to 1990 as mechanization contributed to rising agricultural productivity, reducing the need for labor. Since 1990, employment levels have stabilized. The reduction in self-employed and family labor through 1990 was more rapid than the decline in hired labor. Thus, the proportion of hired workers has increased over time (Castillo and Simnitt 2021).

Specialty crop production typically produces high-value output from relatively small areas of land compared with other crops. Specialty crops are also labor intensive (Astill et al. 2020). A long-term decline in the supply of farm labor in the United States continues to encourage producers to select less labor-intensive crops, invest in labor-saving technologies, and develop strategies to increase labor productivity (Zahniser et al. 2018). The farm sector, especially the specialty crop industry, historically has relied on immigrant farm labor to perform the labor-intensive tasks, such as harvesting. Using the 2015–16 National Agricultural Worker Survey data, Hernandez and Gabbard (2018) estimated that 75% of farmworkers were foreign-born (69% from Mexico) and 49% of farmworkers were unauthorized. Most farmworkers were employed in labor-intensive production: 37% in vegetables, 32% in fruit and nut crops, and 19% in ornamental horticulture. However, the unauthorized immigrant population has declined sharply during the past decade. The recent Pew Research Center estimates show that the population of unauthorized immigrants decreased from its peak of 12.2 million in 2007 to 10.5 million in 2017 as immigration from Mexico slowed down (Passel and Cohn 2019).

In addition to the slow-down of migration from Mexico to the United States, the farm labor workforce is also changing in terms of its mobility. The farmworker population is aging and becoming more settled, traveling less between farms for work (Fan et al. 2015; Martin 2017). The growth of job opportunities in nonagricultural sectors (Taylor et al. 2012; Zahniser et al. 2012) and stringent border control and local immigration enforcement (Ifft and Jodlowski 2016; Kostandini et al. 2014) may also contribute to the reduction in farmworker supply. Two relevant state and local immigration policies that restrict unauthorized immigrants are the 287(g) program and the E-Verify program. The 287(g) program authorizes police to hold individuals for US Immigration and Customs Enforcement (ICE) who have been detained for suspected criminal behavior, whereas the E-Verify program enacted at the state level mandates that employers ascertain an employee’s lawful presence and eligibility to work (Karoly and Perez-Arce 2016). In addition to anecdotes of persistent shortages of farm labor, empirical evidence has shown that these shortages cannot be addressed through higher wages. For example, Hertz and Zahniser (2013) identified localized and crop-specific labor shortages by analyzing wage and employment using multiple data sources. Studies by Martin (2017) and Richards (2018) suggested that the shortage of harvest workers was persistent in California and that stricter immigration policies and labor regulations would increase labor costs by 22% in California. Wei et al. (2019) demonstrated that the rate of substitution between domestic and immigrant labor was low, and little evidence supports that immigrant farmworkers affect the employment of domestic farmworkers.

With ongoing labor shortages, the H-2A program is becoming an increasingly popular source of farmworkers. These shortages are mainly the result of an insufficiently able and willing supply of local and domestic labor (Bampasidou and Salassi 2019). Created in 1952 by the Immigration and Nationality Act, the H-2A program allows farmers to hire nonimmigrant foreign workers to perform seasonal or temporary agricultural work in the United States. The program has expanded tremendously in the past decade. Despite growing participation, farmers often report the application procedures as burdensome and time-consuming, which presents a hurdle for farmers who must find workers in a short window of time. To recruit H-2A workers, growers also incur significant pre-employment costs, such as filing fees, advertising, surety bonds, and travel and housing costs for H2A workers. Roka et al. (2016) estimated that the pre-employment costs associated with H-2A workers were about $2000 per worker. The procedural burden and pre-employment costs may impose a potential hurdle for small and medium-size farms to recruit H-2A workers.

Therefore, we are motivated to investigate what kind of horticultural growers are hiring H-2A workers. Combining the 2014 and 2019 National Green Industry Survey data, we investigated the hiring decision process of H-2A workers by horticultural firms. To date, only a few studies have considered the labor hiring decisions in the horticultural industry [e.g., Bellenger et al. (2008), Krahe and Campbell (2016), Posadas et al. (2014)]. Using an Alabama horticulture industry survey data, Bellenger et al. (2008) showed government regulation is more relevant to decisions hiring migrant workers relative to local workers. Posadas et al. (2014) examined factors affecting hiring decisions on permanent and part-time workers and found southern nursery and greenhouse growers with higher annual sales employ more permanent and part-time workers whereas larger firms hire fewer permanent workers than other nursery and greenhouse growers on a production-per-acre basis. Krahe and Campbell (2016) suggested technologies may have a complementary effect for full- and part-time labor employed by green industry firms.

Materials and Methods

Data.

To investigate the employment of H-2A workers in the ornamental horticulture industry, we use information collected primarily from the 2014 and 2019 National Green Industry Survey (Hodges et al. 2015; Khachatryan et al. 2020). The Green Industry Research Consortium has conducted the National Green Industry Survey at 5-year intervals beginning in 1989 and has collected information on production, marketing practices, and trade flows for US ornamental plant grower and dealer firms for the previous calendar or fiscal year before the survey year. However, only the most recent two surveys (2014 and 2019) gathered employment information on H-2A workers. After removing dealer firms that are not engaged in greenhouse and nursery production, we were left with a total of 2390 observations, 1532 growers from survey year 2014, and 1398 growers from survey year 2019. As shown in Table 1, average employment reported by survey respondents has actually decreased since 2014. In 2014, an average of 11 full-time workers and eight part-time workers per firm was reported, whereas in 2019, the average employment slightly decreased to eight full-time workers and six part-time workers per firm. Although most surveyed firms reported they did not hire H-2A workers in the two waves of surveys, 41 firms in the 2014 survey and 47 firms in the 2019 firms reported a total of 1137 and 981 H-2A workers respectively. This led to an average of 27 H-2A workers in the 2014 survey year, decreasing to an average of 21 H-2A workers in the 2019 survey among those firms that hired H-2A workers. Compared with the average of 14 H-2A workers per farm across all agricultural sectors (Martin 2017), firms in the nursery and floriculture production industry hire more H-2A workers on average than other sectors of agriculture. [According to Martin (2017), the number of farm jobs certified by the US Department of Labor to be filled by H-2A workers was ∼140,000 jobs on 7500 farms in fiscal year 2015. For every 130 farm jobs certified, the US Department of State issued ∼100 H-2A visas. Based on this information, we can easily compute the number of H-2A visas averaged to each farm as ∼14.]

Table 1.

Number of reported full-time, part-time, and H-2A workers by firms in the National Nursery Survey for 2014 and 2019.

Table 1.

In addition to employment information, the National Green Industry Survey asked questions on a host of information, from production and marketing practices to trade flows. Summary statistics for variables included in the analysis can be found in Table 2. Roughly 50% of the surveyed firms are from the Southeast (28%) and the Northeast (20%) United States. The remaining 52% of survey respondents are scattered among the other six physiographic regions. The top three grown plant categories are shrubs (the category shrubs broadly includes deciduous shade and flowering trees, deciduous shrubs, broad-leaved evergreen shrubs, narrow-leaved evergreen shrubs, and azaleas), annual bedding plants (including both flowering annuals and vegetables, fruit, and herbs), and evergreen trees. Forty-three percent of firms reported having sales revenue generated from shrubs, 29% reported sales revenue from flowering annual bedding plants, and 27% reported sales revenue from bedding plants of vegetables, fruit, and herbs, and evergreen trees. Firms in the two surveys are relatively old and were established more than 30 years ago. The average sales value is $1.84 million per firm. In terms of marketing practices, average reported share of sales to repeat customers is 74% and is 17% for negotiated sales. Plant crops are most likely sold in the firm’s home state, reaching 82% on average. Twenty-one percent of the firms in our sample reported they resell and broker plants to other growers.

Table 2.

Summary statistics from the National Nursery Survey for 2014 and 2019.

Table 2.

The National Green Industry Survey also asked firms about their perceptions regarding factors affecting their business on a 4-point rating scale, with 1 point indicating not important and 4 points indicating very important. With an average rating score of 3.3 and 3.2 points, survey respondents believe market demand and weather uncertainty are the two most important factors affecting their business compared with other factors such as labor (2.68 points), competition/price undercutting (2.51 points), and nonenvironmental government regulations (2.36 points). In addition, firms rated “ability to hire competent hourly employees” (2.54 points) as a more important factor affecting business than “ability to hire competent management” (2.11 points).

The National Green Industry Survey data were then merged with information on the adoption of the 287(g) program and the E-Verify program to control for the potential impacts of specific state-level immigration-related policies on individual firm hiring decisions. As of Nov 2021, ICE had 287(g) Jail Enforcement Model agreements with 66 law enforcement agencies in 19 states—an increase from 32 law enforcement agencies in 16 states in 2015. ICE also had 287(g) Warrant Service Officer agreements with 76 law enforcement agencies in Nov 2021 (US Immigration and Customs Enforcement n.d.). Meanwhile, requiring employers in the public and private sectors to verify employment eligibility through E-Verify has been adopted gradually by states. By 2015, eight states required E-Verify for all employees (some states have exemptions for small businesses). These eight states are Alabama, Arizona, Georgia, Mississippi, North Carolina, South Carolina, Tennessee, and Utah. In addition, Louisiana requires that private employers must either use E-Verify or retain work authorization documents (National Conference of State Legislatures 2015). Based on these data, three state-level policy variables [number of law enforcement agencies that participated in the 287(g) program in 2015 and 2019, as well as a binary variable indicating that a state mandates the use of E-Verify for all business] were generated to reflect the state variations in policy restrictiveness toward unauthorized immigrants.

We further supplemented the 2014 and 2019 National Green Industry Survey information with the Quarterly Census of Employment and Wages data and American Community Survey (ACS) data. To control for general demographic influence about the area where the firms are located, we incorporated the ACS 5-year estimates of total population, total male population, mean household income, and unemployment rate at the five-digit zip code level for 2013 and 2018. In addition, the annual number of firm establishments, total employment, and total wage payment for the nursery and floriculture production industry [North American Industry Classification System (NAICS) 11142] are also included to control for aggregate industry influence.

Choice of models.

To model the hiring decision process of H-2A workers, several functional forms were considered. Because the basic hiring decision is a dichotomous choice of whether to participate in an H-2A program, binary response models such as probit and logit are natural considerations. When the choice concerns the level of program participation, a continuous number of H-2A workers needs to be selected in addition to the binary choice of program participation. A corner solution may arise when some responses pile up at zero, whereas others are strictly positive values. The most straightforward way to model a corner solution is the tobit model (Tobin 1958). The tobit model assumes that zero responses and positive values are generated from the same mechanisms.

One extension to the tobit model is a hurdle model (Cragg 1971), in which different mechanisms are allowed to differentiate the H-2A program participation decision and decision on the number of H-2A workers. For example, some factors may have a positive effect on the participation decision, but a negative effect on the number of H-2A workers to be hired. We used a probit model for the binary participation decision, and a truncated normal regression to predict the positive number of H-2A workers. We applied the Vuong test (Vuong 1989) to choose between the tobit and truncated normal hurdle model alternatives. We preferred the Vuong test over the commonly used likelihood ratio test (Greene 2000) for model selection because of our data structure. As mentioned previously, the National Green Industry Survey data were collected using a stratified random sampling procedure by firm size. When clustering occurs, individual observations are no longer independent, and the “likelihood” for clustered maximum log-likelihood estimations is not a true likelihood for sample distribution (Sribney n.d.). With a Vuong test, the likelihood is calculated for each individual observation. The difference of individual likelihood is then taken between the tobit and truncated normal hurdle models and is tested for which model fit better statistically.

Tobit model.

The level of participation in an H-2A program is indicated by a latent variable yi*. The standard type I tobit specification is defined as
yi=max[yi*, 0]
yi*=xiβ+ui>0 uiN(0,σu2),
where y* is a latent endogenous variable representing an ornamental horticultural firm’s desired number of H-2A workers, yi is the actual observed number of H-2A workers hired, xi is a set of variables that explains the H-2A program participation decision, and β is the corresponding vector of parameters to be estimated. In our study, we included factors such as individual firm business and production characteristics, general demographic information about the area where individual firms are located (five-digit zip code level), and industry-level employment and income of the nursery and floriculture production industry in the regression (Table 3). ui is assumed to be a normally distributed error term with zero mean and variance σu2. Because yi is observed only if a positive number of H-2A workers is desired and is zero otherwise, the probability density function consists of two parts:
f(yi | xi)=[1Φ(xiβσu)]1[yi=0]×[1σuΦ(yixiβσu)]1[yi>0].
Table 3.

H-2A workers hiring decision: results from Tobit model.

Table 3.
The unconditional expectation of the number of H-2A workers is
E(yi|xi)=Φ(xiβσu)xiβ+σuΦ(xiβσu).
For each individual firm i, the associated loglikelihood function can be written as
li=1[yi=0] log[1Φ(xiβσu)]+1[yi>0]log[1σuΦ(yixiβσu)].

Truncated normal hurdle model.

The hurdle model proposed by Cragg (1971) is a natural two-part extension of the type I tobit model, which allows different mechanisms to drive the participation decision and the decision on the number of H-2A workers to be recruited. Let yi be the number of H-2A workers hired by an individual firm i, which can be expressed as a compound function of a binary participation decision variable w and the choice of positive number of H-2A workers yi*:
yi=wiyi*.
When a firm decides to participate in the H-2A program (wi = 1), a positive number yi* = yi>0 is observed. When a firm decides not to participate (wi = 0), then yi = 0, and yi* is not observed. The first-stage binary decision is assumed to follow a probit model:
P(wi=1 | xi)=P(yi>0 | xi)=Φ(xiΥσμ) uiN(0,σμ2).
Furthermore, yi* is assumed to have a truncated normal distribution with parameters that vary freely from those in the probit model:
yi*=xiβ+εi>0 εiN(0,σε2),
where εi given xi has a truncated normal distribution with a lower truncation point – xiβ. Like the tobit model, the positive number of H-2A workers yi* depends on firm characteristics and aggregate-level information in xi that influence the decision of how many workers to hire.
  Because yi = yi* when yi > 0, the density of yi for the truncated part is
f(yi | xi, yi>0)=[Φ(xiβσε)]1×[1σεΦ(yixiβσε)], yi>0.
The unconditional probability density function of yi given xi is then
f(yi | xi)=[1Φ(xiYσμ)]1[yi=0]×{Φ(xiYσμ)[Φ(xiβσε)]1×[1σεΦ(yixiβσε)]}1[yi>0],
where we must multiply f(yi | xi, yi>0) by the probability P(yi>0 | xi) = Φ(xiYσμ) for observations yi > 0.
  The unconditional expectation of the number of H-2A workers is
E(yi |xi) = P(wi = 1 | xi)E(yi | xi,wi = 1)= Φ(xiYσμ)[xiβ + σελ(xiβσε)].
The log-likelihood function for individual firm i can be written as
li = 1[yi = 0] log[1Φ(xiΥσμ)]+1[yi>0]log[Φ(xiΥσμ)]+1[yi>0]{log[Φ(xiβσε)]+log[Φ(yixiβσε)]log(σε)}.

Results

We first conducted a Vuong test to evaluate model performance. The Vuong test result based on individual log likelihood between the tobit model (Table 3) and truncated normal hurdle model (Table 4) indicates that the truncated normal hurdle model performs better and fits our data better (P = 0.000) than the tobit model. This is consistent with previous studies that applied both tobit and hurdle models to estimate participation decisions and levels of participation (e.g., Adusah-Poku and Takeuchi 2019; Ma et al. 2012), and suggested the hurdle model was preferred. We focus our discussion on the direction and magnitude of the marginal effects from the truncated normal hurdle model (Table 4). Coefficients of the truncated normal hurdle model are presented in Supplemental Table S1.

Table 4.

H-2A workers hiring decision: marginal effects from truncated normal hurdle model.

Table 4.

In general, we do not see regional differences in hiring H-2A workers, with insignificant coefficients for regional dummies and aggregate ACS demographic variables at the five-digit zip code level (Tables 3 and 4). The number of 287(g) program enforcement agencies does have a significant effect on firms’ participation in H-2A programs. On average, increasing one enforcement agency in a state in 2019 increased the likelihood of participating in H-2A worker programs by 0.2% points (10% significance level). Given the fact that only 3% of firms in our sample reported using H-2A workers, this is a fairly significant effect.

Firm production and business characteristics play more important roles in the H-2A worker program participation decision than the number of hiring decisions. In particular, increasing the sales value by $1 million will increase the likelihood of firms hiring H-2A workers by 0.1% points (1% significance level), revealing that large firms are major players. The business of the ornamental horticulture industry is diversified, with large-scale industry operators and many family-owned small businesses as well. With a median sales value of $125,000 and a mean sales value of $1.84 million in our sample, the economic significance of this effect is marginal.

Firms with repeat customers are more likely to hire H-2A workers. Increasing repeat sales by 10% will increase the probability of firms hiring H-2A workers by 0.5%. On the other hand, increasing sales in the home state is likely to decrease the probability of firms hiring H-2A workers. A 10% increase in sales in the home state leads to a 0.4% reduction in the likelihood of firms recruiting H-2A workers. In contrast, firms attending more tradeshows are more likely to participate in H-2A worker programs. Each additional tradeshow attended will increase the likelihood of recruiting H-2A workers by 0.1%.

When the hurdle of participation is overcome, firm sales value and marketing practices do not matter significantly. On the other hand, specific plant types produced by individual growers play a different role in the first stage of whether to participant in the H-2A workers program and the number of H-2A workers to hire. For example, firms growing evergreen trees and roses are more likely to participate in the H-2A worker program than other firms. Firms growing annual bedding flowers are hiring more H-2A workers, whereas firms growing flowering potted plants are hiring fewer H-2A workers than other firms. The significant effects of the plant type categories in the second stage should be interpreted with caution. Firm participation in the H-2A program is limited (3%) in our sample, and the observed significant effects could be driven by the plant type categories reported by the 51 specific firms in the second stage. However, considering flowering annual bedding plants and evergreen trees are the top-tier ornamental plant types sold based on sales value (Khachatryan et al. 2020), these results may suggest that the current H-2A workers program is in favor of larger firms that are more likely to overcome the fixed costs of hiring H-2A workers.

In addition, firms’ perceptions about factors affecting their business largely affect their decisions regarding whether to hire H-2A workers. In particular, firms that perceive weather uncertainty as an important factor that affects their business are 0.7% less likely than other firms to consider hiring H-2A workers. On the other hand, firms that perceive government regulation as an important factor [e.g., increasing rating from 2 points (“minor importance”) to 3 points (“important”) on the 4-point Likert scale] are 0.6% more likely than other firms to consider hiring H-2A workers. This finding mirrors the results by Bellenger et al. (2008), who found producers perceiving government regulations as a threat to their industry are less likely than other firms to hire migrant workers.

As expected, industry aggregate variables are strong predictors of firms’ hiring of H-2A workers. Firms are more likely to participate in the H-2A workers program when observing increasing total wage payment, but are less likely to participate when observing increasing employment status in the industry. Increasing 1000 jobs in the nursery and floriculture production industry (NAICS 11142) will reduce H-2A participation likelihood by 2%. On the other hand, increasing total wage payment in the industry by $1 million will increase the probability by 0.1%. Although an increase in total employment in the nursery and floriculture production industry may discourage firms to participate in the H-2A program, and an increase in total wage payment may encourage firms to participate in the H-2A program, they have the little impact on the number of H-2A workers. [As a robustness check, an alternative specification of the tobit and hurdle models with an increased df was run using categorical variables to control for regional differences and plant-type variations. The main results remain consistent between the two specifications (Tables 3 and 4).]

Discussion and Conclusion

The ornamental horticulture industry has a significant impact on the national, regional, and local economy. In 2018–19, the industry directly employed 1.3 million full- and part-time workers (Hall et al. 2020). With labor expenses counting for 46% of nursery production costs (Zahniser et al. 2012), the ornamental horticulture industry relies on workers to do myriad tasks, such as pruning, applying fertilizers, scouting, spraying pesticides, planting, harvesting, packing, and weeding. As a result of the perishable nature of horticultural goods, a skilled and accessible labor supply is imperative for continued industry growth and stability. The decreasing number of workers, followed by increasing wage rates, could be alarming for the economic well-being of the ornamental horticulture industry, which has already experienced diminishing revenue as a result of considerable within-industry consolidation, increased price competition, and relatively weak consumer demand (Madigan 2021), not counting the surge in demand for houseplant categories due to the COVID-19 pandemic. Nonetheless, labor issues in this industry have received little attention.

We sought to explore factors that affect horticultural growers’ decisions on whether to participate in the H-2A program and how many H-2A workers to hire. Based on combined 2014 and 2019 National Green Industry Survey data, the hurdle model results indicate that individual firm characteristics play different roles in the H-2A program participation decision and the decision of how many workers to hire. In particular, firm size measured by sales value has a significant impact on the decision to participant in the H-2A program. Although most US horticultural operations are too small to need hired workers, larger operations have an interest in the changing US immigration policies hindering labor availability. States (including California, Florida, Washington, and Georgia) that produce mainly specialty crops, fruit, and vegetables have shown an increased dependency on the H-2A program. For example, Rutledge and Taylor (2019) reported a recent surge in H-2A use in California. California farmers’ use of the H-2A program has been traditionally low as a result of the bureaucracy and associated costs.

The procedural burden and pre-employment costs (e.g., housing and transporting) of the current H-2A program may have created a hurdle to block small and medium-size firms from recruiting H-2A workers, as most businesses in the ornamental horticultural industry are small and family-owned. As farmworkers are generally becoming harder to find and retain, horticultural growers must adapt to an era of a continuously declining farm labor supply. Many farm growers have already sought alternative production methods, including hiring H-2A workers, switching to a less labor-intensive crop mix, and adopting labor-saving technologies to reduce dependence on an elastic seasonal labor supply. Given the diversified nature of ornamental plants and the common practices of producing a set of crop mixes within the industry, a more viable solution for horticultural growers could be adopting more efficient labor management practices such as hiring H-2A workers. State nursery and greenhouse growers’ associations can play an important role in offering services to streamline the hiring process (e.g., filing for H-2A visas) and educating employers about legal and best practices (e.g., transportation requirement and compliance with housing). These services would largely help small and medium-size growers overcome the participation hurdle.

References Cited

  • Adusah-Poku, F & Takeuchi, K. 2019 Household energy expenditure in Ghana: A double-hurdle model approach World Dev. 117 266 277 https://doi.org/10.1016/j.worlddev.2019.01.018

    • Search Google Scholar
    • Export Citation
  • Astill, G, Perez, A & Thornsbury, S. 2020 Developing automation and mechanization for specialty crops: A review of U.S. Department of Agriculture Programs Administrative Publication no. 082. https://www.ers.usda.gov/publications/pub-details/?pubid=95827. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Bampasidou, M & Salassi, ME. 2019 Trends in U.S. farm labor and H-2A hired labor: Policy and related issues Choices. 34 1 1 6

  • Bellenger, M, Fields, D, Tilt, K & Hite, D. 2008 Producer preferences for migrant labor and the wage, hours, and gross sales effects in Alabama’s horticulture industry HortTechnology. 18 2 301 307 https://doi.org/10.21273/HORTTECH.18.2.301

    • Search Google Scholar
    • Export Citation
  • Castillo, M & Simnitt, S. 2021 Farm labor https://www.ers.usda.gov/topics/farm-economy/farm-labor/. [accessed 11 Oct 2022]

  • Cragg, JG. 1971 Some statistical models for limited dependent variables with application to the demand for durable goods Econometrica. 39 5 829 844

    • Search Google Scholar
    • Export Citation
  • Fan, M, Gabbard, S, Pena, AA & Perloff, JM. 2015 Why do fewer agricultural workers migrate now? Am J Agric Econ. 97 3 665 679 https://doi.org/10.1093/ajae/aau115

    • Search Google Scholar
    • Export Citation
  • Greene, WH. 2000 Econometric analysis 4th ed Prentice Hall Englewood Cliffs, NJ, USA

  • Hall, CR, Hodges, AW, Khachatryan, H & Palma, MA. 2020 Economic contributions of the Green Industry in the United States in 2018 J Environ Hortic. 38 3 73 79 https://doi.org/10.24266/0738-2898-38.3.73

    • Search Google Scholar
    • Export Citation
  • Hernandez, T & Gabbard, S. 2018 Findings from the National Agricultural Workers Survey (NAWS) 2015–2016: A demographic and employment profile of United States workers Research report no. 13. https://www.dol.gov/sites/dolgov/files/ETA/naws/pdfs/NAWS_Research_Report_13.pdf. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Hertz, T & Zahniser, S. 2013 Is there a farm labor shortage? Am J Agric Econ. 95 2 476 481 https://doi.org/10.1093/ajae/aas090

  • Hodges, A, Khachatryan, H, Hall, C & Palma, M. 2015 Production and marketing practices and trade flows in the United States green industry, 2013 Southern Coop. Ser. Bul. 420

    • Search Google Scholar
    • Export Citation
  • Ifft, J & Jodlowski, M. 2016 Is ICE freezing US agriculture? The impact of local immigration enforcement on farm profitability and structure Paper presented at the 2016 Agricultural and Applied Economics Association Annual Meeting Boston, MA 31 Jul–2 Aug 2016

    • Search Google Scholar
    • Export Citation
  • Karoly, LA & Perez-Arce, F. 2016 A cost-benefit framework for analyzing the economic and fiscal impacts of state-level immigration policies https://www.rand.org/pubs/research_reports/RR1397.html. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Khachatryan, H, Hodges, A, Hall, C & Palma, M. 2020 Production and marketing practices and trade flows in the United States Green Industry, 2018 Southern Coop. Ser. Bul. 421

    • Search Google Scholar
    • Export Citation
  • Kostandini, G, Mykerezi, E & Escalante, C. 2014 The impact of immigration enforcement on the U.S. farming sector Am J Agric Econ. 96 1 172 192 https://doi.org/10.1093/ajae/aat081

    • Search Google Scholar
    • Export Citation
  • Krahe, J & Campbell, B. 2016 The impact of technology on labor makeup: A study of the Green Industry HortTechnology. 26 3 351 357 https://doi.org/10.21273/HORTTECH.26.3.351

    • Search Google Scholar
    • Export Citation
  • Ma, S, Swinton, SM, Lupi, F & Jolejole-Foreman, C. 2012 Farmers’ willingness to participate in payment-for-environmental-services programmes J Agric Econ. 63 3 604 626 https://doi.org/10.1111/j.1477-9552.2012.00358.x

    • Search Google Scholar
    • Export Citation
  • Madigan, J. 2021 Plant & flower growing in the US IBIS World Industry Report 11142. IBISWorld Inc. https://www.ibisworld.com

  • Martin, PL. 2017 Immigration and farm labor: From unauthorized to H-2A for some? https://www.migrationpolicy.org/sites/default/files/publications/Martin-ImmigrationAgricultureH2AWorkers-FINAL.pdf. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • National Conference of State Legislatures 2015 State E-Verify action https://www.ncsl.org/research/immigration/state-e-verify-action.aspx. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Passel, JS & Cohn, D. 2019 Mexicans decline to less than half the U.S. unauthorized immigrant population for the first time https://www.pewresearch.org/fact-tank/2019/06/12/us-unauthorized-immigrant-population-2017/. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Posadas, BC, Knight, PR, Coker, CEH, Coker, RY & Langlois, SA. 2014 Hiring preferences of nurseries and greenhouses in U.S. southern states HortTechnology. 24 1 107 117 https://doi.org/10.21273/HORTTECH.24.1.107

    • Search Google Scholar
    • Export Citation
  • Richards, TJ. 2018 Immigration reform and farm labor markets Am J Agric Econ. 100 4 1050 1071 https://doi.org/10.1093/ajae/aay027

  • Roka, FM, Simnitt, S & Farnsworth, D. 2016 Pre-employment costs associated with H-2A agricultural workers and the effects of the ‘60-minute rule’ Int Food Agribusiness Manag Rev. 20 3 335 346 https://doi.org/10.22434/IFAMR2016.0033

    • Search Google Scholar
    • Export Citation
  • Rutledge, Z & Taylor, JE. 2019 California farmers change production practices as the farm labor supply declines https://www.zachrutledge.com/uploads/1/2/5/6/125679559/are_update_article_final_draft.pdf. [accessed 1 Dec 2022]

    • Search Google Scholar
    • Export Citation
  • Sribney, W n.d. Why should I not do a likelihood-ratio test after an ML estimation (e.g., logit, probit) with clustering or pweights? http://www.stata.com/support/faqs/stat/lrtest.html. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Taylor, JE, Charlton, D & Yúnez-Naude, A. 2012 The end of farm labor abundance Appl Econ Perspec Policy. 34 4 587 598 https://doi.org/10.1093/aepp/pps036

    • Search Google Scholar
    • Export Citation
  • Tobin, J. 1958 Estimation of relationships for limited dependent variables Econometrica. 26 1 24 36

  • US Bureau of Labor Statistics 2022 Employment and Wages, Annual Averages Bulletins, 2001-forward https://www.bls.gov/cew/publications/employment-and-wages-annual-averages/

    • Search Google Scholar
    • Export Citation
  • US Census Bureau 2022 American Community Survey 5-Year Data (2009–2021) https://www.census.gov/data/developers/data-sets/acs-5year.html

  • US Immigration and Customs Enforcement n.d Delegation of Immigration Authority Section 287(g) Immigration and Nationality Act https://www.ice.gov/identify-and-arrest/287g. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Vuong, QH. 1989 Likelihood ratio tests for model selection and non-nested hypotheses Econometrica. 57 2 307 333

  • Wei, X, Guan, Z, Onel, G & Roka, F. 2019 Substitution between immigrant and native farmworkers in the United States: Does legal status matter? IZA J Dev Migr. 10 3 https://doi.org/10.2478/izajodm-2019-0007

    • Search Google Scholar
    • Export Citation
  • Zahniser, S, Hertz, T, Dixon, P & Rimmer, M. 2012 The potential impact of changes in immigration policy on U.S. agriculture and the market for hired farm labor: A simulation analysis U.S. Dept. Agr. Econ. Res. Ser, Econ. Res. Rpt. no. 135. https://www.ers.usda.gov/webdocs/publications/44981/err-135.pdf. [accessed 10 Jun 2021]

    • Search Google Scholar
    • Export Citation
  • Zahniser, S, Taylor, JE, Hertz, T & Charlton, D. 2018 Farm labor markets in the United States and Mexico pose challenges for U.S. agriculture U.S. Dept. Agr. Econ. Res. Ser. Info. Bul. 201. https://www.ers.usda.gov/webdocs/publications/90832/eib-201.pdf?v=7176.5. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation

Supplemental Table S1.

H-2A workers hiring decision: estimated coefficients from the truncated normal hurdle mode.

Supplemental Table S1.
  • Adusah-Poku, F & Takeuchi, K. 2019 Household energy expenditure in Ghana: A double-hurdle model approach World Dev. 117 266 277 https://doi.org/10.1016/j.worlddev.2019.01.018

    • Search Google Scholar
    • Export Citation
  • Astill, G, Perez, A & Thornsbury, S. 2020 Developing automation and mechanization for specialty crops: A review of U.S. Department of Agriculture Programs Administrative Publication no. 082. https://www.ers.usda.gov/publications/pub-details/?pubid=95827. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Bampasidou, M & Salassi, ME. 2019 Trends in U.S. farm labor and H-2A hired labor: Policy and related issues Choices. 34 1 1 6

  • Bellenger, M, Fields, D, Tilt, K & Hite, D. 2008 Producer preferences for migrant labor and the wage, hours, and gross sales effects in Alabama’s horticulture industry HortTechnology. 18 2 301 307 https://doi.org/10.21273/HORTTECH.18.2.301

    • Search Google Scholar
    • Export Citation
  • Castillo, M & Simnitt, S. 2021 Farm labor https://www.ers.usda.gov/topics/farm-economy/farm-labor/. [accessed 11 Oct 2022]

  • Cragg, JG. 1971 Some statistical models for limited dependent variables with application to the demand for durable goods Econometrica. 39 5 829 844

    • Search Google Scholar
    • Export Citation
  • Fan, M, Gabbard, S, Pena, AA & Perloff, JM. 2015 Why do fewer agricultural workers migrate now? Am J Agric Econ. 97 3 665 679 https://doi.org/10.1093/ajae/aau115

    • Search Google Scholar
    • Export Citation
  • Greene, WH. 2000 Econometric analysis 4th ed Prentice Hall Englewood Cliffs, NJ, USA

  • Hall, CR, Hodges, AW, Khachatryan, H & Palma, MA. 2020 Economic contributions of the Green Industry in the United States in 2018 J Environ Hortic. 38 3 73 79 https://doi.org/10.24266/0738-2898-38.3.73

    • Search Google Scholar
    • Export Citation
  • Hernandez, T & Gabbard, S. 2018 Findings from the National Agricultural Workers Survey (NAWS) 2015–2016: A demographic and employment profile of United States workers Research report no. 13. https://www.dol.gov/sites/dolgov/files/ETA/naws/pdfs/NAWS_Research_Report_13.pdf. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Hertz, T & Zahniser, S. 2013 Is there a farm labor shortage? Am J Agric Econ. 95 2 476 481 https://doi.org/10.1093/ajae/aas090

  • Hodges, A, Khachatryan, H, Hall, C & Palma, M. 2015 Production and marketing practices and trade flows in the United States green industry, 2013 Southern Coop. Ser. Bul. 420

    • Search Google Scholar
    • Export Citation
  • Ifft, J & Jodlowski, M. 2016 Is ICE freezing US agriculture? The impact of local immigration enforcement on farm profitability and structure Paper presented at the 2016 Agricultural and Applied Economics Association Annual Meeting Boston, MA 31 Jul–2 Aug 2016

    • Search Google Scholar
    • Export Citation
  • Karoly, LA & Perez-Arce, F. 2016 A cost-benefit framework for analyzing the economic and fiscal impacts of state-level immigration policies https://www.rand.org/pubs/research_reports/RR1397.html. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Khachatryan, H, Hodges, A, Hall, C & Palma, M. 2020 Production and marketing practices and trade flows in the United States Green Industry, 2018 Southern Coop. Ser. Bul. 421

    • Search Google Scholar
    • Export Citation
  • Kostandini, G, Mykerezi, E & Escalante, C. 2014 The impact of immigration enforcement on the U.S. farming sector Am J Agric Econ. 96 1 172 192 https://doi.org/10.1093/ajae/aat081

    • Search Google Scholar
    • Export Citation
  • Krahe, J & Campbell, B. 2016 The impact of technology on labor makeup: A study of the Green Industry HortTechnology. 26 3 351 357 https://doi.org/10.21273/HORTTECH.26.3.351

    • Search Google Scholar
    • Export Citation
  • Ma, S, Swinton, SM, Lupi, F & Jolejole-Foreman, C. 2012 Farmers’ willingness to participate in payment-for-environmental-services programmes J Agric Econ. 63 3 604 626 https://doi.org/10.1111/j.1477-9552.2012.00358.x

    • Search Google Scholar
    • Export Citation
  • Madigan, J. 2021 Plant & flower growing in the US IBIS World Industry Report 11142. IBISWorld Inc. https://www.ibisworld.com

  • Martin, PL. 2017 Immigration and farm labor: From unauthorized to H-2A for some? https://www.migrationpolicy.org/sites/default/files/publications/Martin-ImmigrationAgricultureH2AWorkers-FINAL.pdf. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • National Conference of State Legislatures 2015 State E-Verify action https://www.ncsl.org/research/immigration/state-e-verify-action.aspx. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Passel, JS & Cohn, D. 2019 Mexicans decline to less than half the U.S. unauthorized immigrant population for the first time https://www.pewresearch.org/fact-tank/2019/06/12/us-unauthorized-immigrant-population-2017/. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
  • Posadas, BC, Knight, PR, Coker, CEH, Coker, RY & Langlois, SA. 2014 Hiring preferences of nurseries and greenhouses in U.S. southern states HortTechnology. 24 1 107 117 https://doi.org/10.21273/HORTTECH.24.1.107

    • Search Google Scholar
    • Export Citation
  • Richards, TJ. 2018 Immigration reform and farm labor markets Am J Agric Econ. 100 4 1050 1071 https://doi.org/10.1093/ajae/aay027

  • Roka, FM, Simnitt, S & Farnsworth, D. 2016 Pre-employment costs associated with H-2A agricultural workers and the effects of the ‘60-minute rule’ Int Food Agribusiness Manag Rev. 20 3 335 346 https://doi.org/10.22434/IFAMR2016.0033

    • Search Google Scholar
    • Export Citation
  • Rutledge, Z & Taylor, JE. 2019 California farmers change production practices as the farm labor supply declines https://www.zachrutledge.com/uploads/1/2/5/6/125679559/are_update_article_final_draft.pdf. [accessed 1 Dec 2022]

    • Search Google Scholar
    • Export Citation
  • Sribney, W n.d. Why should I not do a likelihood-ratio test after an ML estimation (e.g., logit, probit) with clustering or pweights? http://www.stata.com/support/faqs/stat/lrtest.html. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Taylor, JE, Charlton, D & Yúnez-Naude, A. 2012 The end of farm labor abundance Appl Econ Perspec Policy. 34 4 587 598 https://doi.org/10.1093/aepp/pps036

    • Search Google Scholar
    • Export Citation
  • Tobin, J. 1958 Estimation of relationships for limited dependent variables Econometrica. 26 1 24 36

  • US Bureau of Labor Statistics 2022 Employment and Wages, Annual Averages Bulletins, 2001-forward https://www.bls.gov/cew/publications/employment-and-wages-annual-averages/

    • Search Google Scholar
    • Export Citation
  • US Census Bureau 2022 American Community Survey 5-Year Data (2009–2021) https://www.census.gov/data/developers/data-sets/acs-5year.html

  • US Immigration and Customs Enforcement n.d Delegation of Immigration Authority Section 287(g) Immigration and Nationality Act https://www.ice.gov/identify-and-arrest/287g. [accessed 12 Jan 2023]

    • Search Google Scholar
    • Export Citation
  • Vuong, QH. 1989 Likelihood ratio tests for model selection and non-nested hypotheses Econometrica. 57 2 307 333

  • Wei, X, Guan, Z, Onel, G & Roka, F. 2019 Substitution between immigrant and native farmworkers in the United States: Does legal status matter? IZA J Dev Migr. 10 3 https://doi.org/10.2478/izajodm-2019-0007

    • Search Google Scholar
    • Export Citation
  • Zahniser, S, Hertz, T, Dixon, P & Rimmer, M. 2012 The potential impact of changes in immigration policy on U.S. agriculture and the market for hired farm labor: A simulation analysis U.S. Dept. Agr. Econ. Res. Ser, Econ. Res. Rpt. no. 135. https://www.ers.usda.gov/webdocs/publications/44981/err-135.pdf. [accessed 10 Jun 2021]

    • Search Google Scholar
    • Export Citation
  • Zahniser, S, Taylor, JE, Hertz, T & Charlton, D. 2018 Farm labor markets in the United States and Mexico pose challenges for U.S. agriculture U.S. Dept. Agr. Econ. Res. Ser. Info. Bul. 201. https://www.ers.usda.gov/webdocs/publications/90832/eib-201.pdf?v=7176.5. [accessed 11 Oct 2022]

    • Search Google Scholar
    • Export Citation
Xuan Wei Food and Resource Economics Department and Mid-Florida Research and Education Center, University of Florida, Apopka, FL 32703, USA

Search for other papers by Xuan Wei in
Google Scholar
Close
,
Benjamin L. Campbell Department of Agricultural and Applied Economics, University of Georgia, Athens, GA 30602, USA

Search for other papers by Benjamin L. Campbell in
Google Scholar
Close
,
Hayk Khachatryan Food and Resource Economics Department and Mid-Florida Research and Education Center, University of Florida, Apopka, FL 32703, USA

Search for other papers by Hayk Khachatryan in
Google Scholar
Close
, and
Robin G. Brumfield Department of Agricultural, Food & Resource Economics, Rutgers, The State University of New Jersey, New Brunswick, NJ 08901, USA

Search for other papers by Robin G. Brumfield in
Google Scholar
Close

Contributor Notes

This project was supported by a grant from The Horticulture Research Institute (project no. 5927314), with cost sharing provided by the University of Florida and Texas A&M University.

X.W. and B.L.C. are the corresponding authors. E-mail: wei.xuan@ufl.edu and ben.campbell@uga.edu.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 885 461 57
PDF Downloads 523 323 56
Save