Top-down Farm Worker Estimates

 Click here to download this blog post as a PDF file

May 13, 2021

How many farm workers are there in each state? This is a surprisingly difficult question to answer. A farm worker is usually defined as a person who works for wages on a farm, whether for one day or for 365 days. There are many types of farm workers, from field workers who harvest crops for a few weeks or months to equipment operators and irrigators to supervisors, mechanics, and those who work with animals year-round.

Bottom up. The federal government distributes at least $1 billion a year to support migrant and seasonal farm workers (MSFWs) and their dependents, most via the Big 4 programs of Migrant Education, Migrant Heath, Migrant Head Start, and the National Farmworker Jobs Program. These programs serve a fraction of the MSFWs that are believed to be eligible for services. However, each program has a different definition of who is eligible for services, and each uses a different methodology to estimate the distribution of eligible MSFWs in order to distribute federal funds across states.

Some programs use bottom-up surveys to estimate the number of eligible MSFWs, which means that they begin with MSFWs and their children who receive services or are identified as eligible for services and multiply by a factor to account for other MSFWs. However, such studies lack a reliable factor for multiplying.

Another bottom-up approach is to combine data from MSFW programs and other sources to estimate MSFWs and dependents who satisfy eligibility criteria. The Lillesand estimate of MSFWs and dependents by state prepared for the Legal Services Corporation in 1977 used mixed methods to estimate five million MSFWs (including dependents), of whom 1.6 million were migrants and 3.4 million were seasonal farm workers and their dependents.

A bottom-up study estimated 5 million MSFWs and dependents in 1977

Other bottom-up studies begin with the acreage of various commodities by county, estimate the hours per acre needed to produce and harvest them, and divide by average hours worked by each worker to estimate the number of farm workers by commodity and state. The result was an estimated 732,109 MSFWs in California in 2000, plus 570,000 dependents.

A 2000 bottom-up study estimated 732,000 MSFWs in CA in 2000

Bottom-up studies were examined in 1988 and 1994 books and rejected as a reliable estimation method. Bottom-up studies that begin with MSFWs who receive services have difficulty developing reliable multipliers to move from those being served to the universe of MSFWs who are eligible for services. Studies that begin with acreage and hours worked have difficulty accounting for the variance in production methods, and find it very hard to develop a reliable indicator of average hours worked by those employed in each commodity.

Top-down. Top-down farm worker estimation strategies begin with the most reliable data available, often farm operator expenses incurred by employers for workers that are hired directly and for contract labor brought to farms. These labor expense data from the Census of Agriculture (most recent is 2017) can be converted into hours worked by dividing them by the average hourly earnings of US crop workers ($11.99 in 2017 from the NAWS) and US livestock workers ($12.38 from the FLS).

Hours worked, in turn, can be converted into farm workers in each state by dividing them by the average number of hours worked by crop workers (1,388 hours a year from the NAWS) and livestock workers (2,030 hours a year from FLS weekly hours worked times 50 weeks).

This top-down method generates an estimate of 1.6 million US crop and 510,000 US livestock workers in 2017, for a total of almost 2.2 million US farm workers.

These farm worker estimates highlight three major items, the dominance of crop workers, the concentration of farm workers in a few states, and differences in the distribution of crop and livestock workers across states. First, three fourths of farm workers are employed by crop farms. Across states, almost 95 percent of farm workers were employed on crop farms in Hawaii, versus a third in Wyoming. Six states, including WI and OK, had fewer than half of their farm workers employed on crop farms, while nine states, including WA, CA, and FL, had more than 85 percent of farm workers employed on crop farms.

Second, farm workers are concentrated in a few states. California’s estimated 537,000 farm workers in 2017 are a quarter of the US total, followed by 125,000 or almost six percent in WA, 117,000 or over five percent in FL, and 103,000 or almost five percent in TX. The four states that each had over 100,000 farm workers had 41 percent of US farm workers in 2017.

Six states each had 50,000 to 100,000 farm workers, beginning with NC with 92,000, MI with 70,000, WI and OR with 60,000 each, MN with 55,000, and NY and PA with 50,000 each. These six states had a total of 440,000 or 20 percent of US farm workers.

Another 15 states had 25,000 to 50,000 farm workers in 2017, led by IA with 48,000, IL with 44,000, and GA, KY, ID, VA, and AZ with 40,000 each. Three states had about 35,000 farm workers each, OH, NE, and TN, and KS, IN, MO, and CO had 30,000 to 33,000 farm workers each. These 15 states had a total of 525,000 farm workers or 24 percent of US farm workers in 2017.

Third, there are differences in shares of crop and livestock workers across states. The 13 states that each had at least two percent of US farm workers had two-thirds of all farm workers in 2017, including 69 percent of crop workers and 53 percent of livestock workers, that is, crop workers are more concentrated in relatively few states than livestock workers.

CA stands out as having 30 percent of US crop workers, four times more crop workers than any other state. TX and CA each have about 10 percent of livestock workers.

CA had 30% of US crop workers in 2017; TX had 10% of US livestock workers

Changes. Applying the same top-down methodology to COA labor expenses in 2002, 2007, and 2012 shows remarkable stability in the total number of farm workers, about two million, and the distribution of farm workers across states, including a quarter in California. There were 12 to 16 states that each had two percent or more of all US farm workers, and collectively they had 65 to 70 percent of all farm workers.

California’s share of US farm workers rose from 24 to 29 percent between 2002 and 2007 and then stabilized at 25 percent. WA’s share of US farm workers rose from four percent in 2007 to almost six percent in 2012 and 2017, while Florida’s share fell from over eight percent in 2002 to five percent after 2012. Similarly, the share of farm workers in TX fell from over six percent to less than five percent between 2002 and 2017.

The falling share of farm workers in FL and TX, and the rising share in WA and NC, is clear in the move from the yellow farm worker shares of 2002 to the blue shares of 2017. The share of US farm workers in MI and WI rose, the share in OR fell, and the share in MN and NY rose slightly.

The share of farm workers in FL and TX fell between 2002 and 2017, while the share in WA and NC rose

Top 10 states: shares of US farm workers, 2002, 2007, 2012, and 2017

The National Center for Farmworker Health relies on a variation of the top-down method that divides COA labor expenses for directly hired workers by the number of jobs reported by farmers who hire workers directly. Farmers reported almost $32 billion in expenses for 2.4 million directly hired workers in the 2017 COA, making the average cost per worker or job $13,120.

NCFH divided the $7.6 billion in contract labor expenses by the same $13,120 to estimate an additional 580,000 jobs for contract workers, for a total of almost 2.9 million jobs.

NCFH estimates of direct and contract workers are available by county. Fresno county in 2017 had 38,000 directly hired workers and 34,000 contract workers using this method, while Yakima county had 54,000 directly hired workers and 5,000 contract workers (http://www.ncfh.org/number-of-ag-workers.html).

The COA counts jobs rather than workers, so that one worker employed and reported by two farms is counted twice. COA labor expenses include payments to relatives of the owner and for all types of workers, and include the cost of payroll taxes and employee benefits.

Perspective. During the 1960s, the federal government launched programs to help MSFWs and their children to prepare for nonfarm jobs at a time when labor-saving mechanization was expected to reduce the number of farm jobs. Migrants who moved from one state to another were considered to be a special federal responsibility because many states had residency requirements that limited the eligibility of migrant farm workers to safety net programs. These residency requirements were declared unconstitutional by the US Supreme Court in Shapiro v Thompson in 1969, but the migrant programs expanded to cover both migrant and seasonal workers and continued.

MSFW program administrators commissioned studies of the number of eligible MSFWs to justify requests for funding at a time when manual farm work was believed to be shrinking rapidly. For example, a major DOL-funded study predicted in 1970 that “the farm labor force of the future will consist of a core of skilled and semi-skilled men operating farm machinery supplemented seasonally by unskilled manual laborers, mostly local housewives and school-age youths.” (Cargill and Rossmiller, 1970, p18).

The sense that MSFWs would soon disappear justified the lack of effort to develop reliable methods to estimate the number and distribution of MSFWs. Even as MSFW employment rose during the 1980s and 1990s with expanded production of fresh fruits and vegetables, many programs continued to use ad hoc methods to estimate those eligible for services. Instead of bricks building a stronger wall, the ad hoc surveys used a wide variety of methods, so that most new studies reviewed and rejected previous methodologies.

Top-down strategies have become dominant in the 21st century. They begin with reliable data, make adjustments that are transparent, and can be updated as new data become available. They show that the US has about 2.2 million farm workers, including almost three fourths employed on crop farms and a quarter in California.

Cargill, Burton and George Rossmiller. 1970. Fruit and Vegetable Harvest Mechanization. Policy Implications. MSU.

Martin, Philip and David Martin. 1994. The Endless Quest: Helping America's Farm Workers. Westview Press

Martin, Philip L. 1988. Harvest of Confusion: Migrant Workers in U.S. Agriculture. Boulder: Co. Westview Press.

NCFH. Agricultural Worker Estimates. 2017

Top-down farm worker estimates, 2017

Subscribe via Email

Click here to subscribe to Rural Migration News via email.