The US Bureau of Labor Statistics computes the costs of hiring workers. In March 1999, the most recent data available, the total cost of employing workers in the US private sector was $19 an hour, including $13.87 an hour in wages and salaries (73 percent) and $5.13 an hour in benefits (27 percent).
Fringe benefits included $1.65 an hour or nine percent of total compensation for legally required payments for social security, unemployment insurance and workers' compensation. This means that voluntary benefits were 18 percent of total compensation, including $1.20 an hour or six percent for paid leave (vacation and holiday pay) and $1.13 for health and other insurance. Retirement and savings benefits added $0.57 an hour or three percent to total compensation, and supplemental pay added $0.55 or three percent of total compensation.
Benefits averaged 27 percent of total compensation between 1986 and 1999- the benefit share of total compensation peaked at 29 percent in 1993-94 when social security and health insurance contributions rose. Legally mandated benefits averaged nine percent over this period, meaning that voluntary employer-provided benefits averaged 18 percent of total compensation. U.S. Bureau of Labor Statistics. 2000. Employer Costs for Employee Compensation, 1986-99. March. For more information: http://stats.bls.gov/special.requests/ocwc/oclt/ect/ecbl0013.pdf
Social Security. The Social Security Administration issues numbers to legal US residents in order to track their earnings for the purpose of determining their Social Security benefits; legal nonimmigrants in the US are issued SSNs if they are authorized to work. In California, the Department of Motor Vehicles asks for a SSN from applicants for a driver's license or a "denial" letter from the Social Security Administration stating that the individual is not eligible for the SSN.
The Social Security Administration "rejects" magnetic media filings by employers of W-2 earnings reports when 50 percent or more of the earnings reports have a name-number mismatch, or the SSN was never issued by the SSA. The Social Security Administration says that about 65 percent of mismatch letters are issued in the seven states with the most illegal foreigners. SSA accepts Social Security payments made on behalf of individuals with name-SSN mismatches, putting the taxes into a "suspense file" that in summer 2000 contained $292 billion.
MHS Estimates. The Migrant Health Program (MHP), through the National Center for Farmworker Health, commissioned estimates of the number of migrant, seasonal and nonworking dependents who accompany these workers in ten states: Arkansas, California, Florida, Louisiana, Maryland, Mississippi, North Carolina, Oklahoma, Texas and Washington.
To be eligible for Migrant Health Program services, workers must have their "principal employment" in agriculture, and be or have been employed on a seasonal basis within the past 24 months. A migrant is a person who establishes "a temporary abode" in order to be employed in qualifying "agriculture." Agriculture, for Migrant Health purposes, includes only crops, not livestock, and embraces processing, packing and similar activities only if "performed by a farmer or on a farm incident to or in conjunction with "primary growing or harvesting activity."
Estimates of the number of eligible workers and dependents are made for the purpose of allocating funds, but an individual's eligibility for MHS services is determined by clinic-intake workers when workers and their dependents seek services. Clinics, which often serve farm workers as well as other low-income patients, typically do not call farm employers or requireW-2 earnings statements to verify worker assertions about qualifying farm employment.
The NCFH contracted with Alice Larson to develop county-level estimates. She used a "demand-for-labor" approach to estimate employment for four major MHS-covered commodities: crops, nursery/greenhouses, food processing and reforestation. Most MSFWs are employed in crops, and the estimates for crops in most cases were made by: (1) taking the acres of labor-intensive crops in each county from the 1997 Census of Agriculture; (2) multiplying acres by the estimated hours of hired work per acre used to produce these crops (from a variety of sources, including farm advisors); and then (3) translating total hours of farm work per crop into days of farm work by assuming that: (a) all hours of work in these commodities were done by MSFWs; and (b) workers average 7.7 hours a day.
For example, if fresh tomato production requires 200 hours per acre, and there were 40,000 acres of tomatoes in Florida in 1997, then a total eight million hours of labor are estimated to be needed in tomato production. If workers average 7.7 hours a day, and the peak (harvest) season is 141 days, the average worker would be employed in tomatoes for 1,086 hours (141 days x 7.7 hours). The number of workers employed in tomatoes is thus eight million divided by 1,086 or 7,400.
The two critical variables in estimating the number of workers using this methodology are: (1) the hours-per-acre estimate; and (2) the estimated length of the peak (harvest) season. If hours-per-acre increase, so does the number of workers. However, longer harvest seasons reduce the number of workers.
The interaction between hours per acre and length of the peak season can be made clear by comparing North Carolina and Florida. North Carolina had 1,900 acres of tomatoes in the 1997 Census of Agriculture (COA), and used an estimated 267 hours per acre to produce them, or a total 507,000 hours. The peak tomato season in North Carolina was assumed to be 45 days, so that an average worker would be employed 45 x 7.7 = 346 hours. Dividing 507,000 by 346 means a peak 1,466 tomato workers. Even though North Carolina has five percent as many acres of tomatoes as Florida, it is estimated to have 20 percent as many tomato pickers, primarily because of the shorter North Carolina season.
Larson's spreadsheet procedure estimated that there were 185,000 MSFWs in Florida in 1997, including 114,000 migrants and 71,000 seasonal workers, with another 42,000 nonfarm workers in migrant worker households and 42,000 nonfarm workers in seasonal worker households. By county, about 20,000 MSFWs were in Palm Beach county, followed by 16,600 in Hillsborough; 13,000 in Dade; and 12,000 each in Polk, Collier and Hendry.
Evaluation. The Larson estimates are a very useful first step and would be even more useful if the spreadsheets on which they are based were placed on the Internet so that errors and omissions would be easier to spot and correct. The reports for the 10 states are expected to be available via http://www.ncfh.org/
However, the published estimates seem to produce too many farm workers and too high a percentage of migrant workers. No other recent study: (1) ignored year-round or non-seasonal workers, who are the majority of workers hired directly by farm operators who report employment in the Census of Agriculture; or (2) put the percentage of migrants among MSFWS at 45 percent or more in many states, for example, MSFWs were estimated to be 62 percent migrants and 38 percent seasonals in Florida, and 46 percent migrants and 54 percent seasonals in California.
For example, if year-round workers contribute 30 to 40 percent of the hours of work in a commodity, then Larson's MSFW estimates will be too high. It is not clear how the distinction between migrants and seasonals was made. Most statistical studies of farm workers in California report that 10 to 20 percent are follow-the-crop migrants, meaning that they have two or more farm jobs far enough apart that they may have to establish a temporary US home away from their primary US home to do farm work. If the workers who arrive from Mexico and stay in one US area to do farm work are considered migrants, the percentage rises to 40 to 50 percent, even though most of these Mexican commuters do not move around the US to do farm work.
The 1997 Census of Agriculture reported that 12,200 Florida farms had a total $925 million in farm labor expenditures. Some 7,500 farms- sometimes the same farm that had farm labor expenditures- had another $477 million in contract labor expenses. Some 6,500 farms reported hiring 50,100 workers who were expected to be employed 150 or more days on that farm; 10,400 farms- sometimes the same farm- reported hiring 74,900 workers expected to be employed for less than 150 days on that farm. These COA data suggest that the total number of workers hired directly- even assuming no double counting, as would occur when the same individual was hired by two farmers, was 125,000.
MHP does not cover livestock, which is a relatively small sector in Florida because of hot weather- only 11,500 workers were expected to be employed on livestock farms for less than 150 days in 1997. If livestock workers are subtracted, the COA suggests about 64,000 less than 150-day jobs on Florida crop farms.
Only $9 million of the contract labor expenses were in livestock; about $468 million were on crop farms. Contract labor expenses were translated into seasonal workers by assuming that 18 to 36 percent of the contract labor expenses were FLC overhead to cover payroll taxes and profits, and the remainder of the contract labor expenses were wages paid to seasonal workers. For example, at 36 percent overhead, seasonal worker wages were 0.64 x 468 = $300 million. These estimated contract labor expense wages paid to workers were divided by $6.76 an hour (average hourly earnings in 1997) to estimate that 44 million hours of work were done for FLCs on Florida farms. The average seasonal worker was assumed to work 962 hours, so that 44 million hours translates into 46,223 workers.
Thus, 63,400 less-than-150 day workers hired directly by crop farmers in the 1997 Census, plus 46,223 workers hired by FLCs at a 36 percent overhead rate, equals 109,000 MSFWs. One cross-check of this estimate comes from data on Unemployment Insurance-covered employment and wages. In 1997, BLS reported that there were an average 154,000 workers employed by 12,700 establishments in Agriculture, Forestry, and Fishing in Florida (157,000 were employed by 13,300 establishments in 1998).
Because of worker turnover, more individuals were likely employed in Florida agriculture, but UI-covered employment also includes hired managers on corporate farms as well as their office and other nonfarm employees. All UI-covered employees in agricultural SIC codes had $2.5 billion in UI covered wages in 1997, and $2.7 billion in UI covered wages in 1998; average weekly wages were $320 in 1997, with average annual earnings of $16,700. For more information: http://stats.bls.gov/cewhome.htm)