An extremely large business with large acreage and production value, for example, may have been classed in the small or medium groups if it made extensive use of fles. We acknowledge this problem in stratifying the population for sampling. and we avoid it in the presentation of findings by using other measures of size–total labor expense, production value, acres–as appropriate.The first mailing to businesses in our survey sample was made during late October 1992. An envelope containing a cover letter, final version of the questionnaire, administrative control sheet, summary request postcard, and return envelope was sent to each of the 2,500 selected farm owners or operators. The Postal Service promptly returned more than one-hundred of the envelopes with invalid or nonexistent addresses, thus reducing the effective sample size. Usable returns from farmers began to arrive within a few days and continued to accumulate at a substantial rate for one month. The second mailing, a simple reminder postcard, went out to all non-respondents during the first week of December. By the end of January 1993, a lotol of 586 completed and usable questionnaires had been returned.For every questionnaire recipient who went to the trouble of telling us that he or she was not or might not be part of the population, there were likely others in the same situation who did not bother to communicate. The response rate after the first two mailings did not measure up to expectations based on long experience of the Survey Research Center in similar studies. The sheer length of the survey instrument was considered partly but not solely responsible. We therefore conducted an interim study to assess the extent to which “non-response” was from such people who should not have been in the sample to begin with, and to identify other factors discouraging response that could be modified to improve the third mailing.
The most currently available EDD employment and payroll data on 200 randomly selected non-respondents were acquired and examined,growers solutions and they indicated that 22 of these farms were not currently in business. A sub-sample’ of 20 was then chosen for brief phone interviews, which yielded no single’, clear explanation but did bear out that instrument length and personal time constraints were important factors to recipients. For the third mailing, in March 1993, we devised a short version of the questionnaire one-half the size of the original and excluding items most burdensome to complete. The short form was thus a subset of the long; it contained no new items. The full original questionnaire was·sent to one-third of non-respondents remaining in the sample and the short form to two-thirds. By early May the Postal Service had returned as undeliverable a total of 151 of the envelopes initially mailed to thc 2,500 businesses selected to the sample. Another 89 <3.6 percent of the selected businesses informed us by mail or phone call that they wcre no longer operating, had never been in business, or were not in a farm business. The valid population list thus numbered 2,260, and we heard formally from 955 . There were 3D outright refusals , one farmer regretfully and regrettably “unable to participate” due to major illness, and a total of 924 complete and usable responses before a May 7, 1993, cut-off date. An additional 19completed questionnaires–received after May 7, and one received earlier but not representing what we construe to be a “farm business,” are not included in the data file analyzed for this report. Written or called-in comments from questionnaire recipients who did not participate in the survey were recorded verbatim. These remarks, several quite candid, are grouped by reason for non-participation and presented in Appendix3. The completion rate varies little across payroll size based on VI file data, from a low of 38 percent in the 50-75th percentile group to a high of 46 percent in the 95-99th percentile group. Smaller operators up to the 50th percentile, while close to the sample average in questionnaire completion rate, were markedly more likely to report that they had left farm business.
Survey participation varied more strongly as a function of SIC code. Dairy operators had the lowest response rate among commodity-based SIC groups, and producers of nuts , citrus fruit , and other fruit the highest. Amazingly, response rates on the third mailing were Virtually the same for farmers receiving the original full-length questionnaire as for those who were asked to complete the shortened version . It is difficult to figure.Data from usable responses were entered and verified In a specialized database program into which logical sequences and consistency checks had been built.Findings have been aggregated and are presented in this report mostly as cross-tabulations for the sample as a whole and for farm size, region,l, and commodity groupings. More sophisticated multivariate techniques were used to test for connections between respective farm characteristics and employment practices. Significant relationships are noted in the text of sections D·G. The measure of size used in most tables, particularly those referring to management of the farm’s own employees, is direct payroll as reported on the questionnaire. Total labor expense, including direct payroll plus payments to contractors and other labor service providers, is used in presentation of findings more logically related to size of the entire farm operation. Geographical classification is based on the county in which the respondent farm reports prodcuing its greatest revenue, not on its address of record in the VI file that defined the survey population. The addresses listed in the EDD record for many operators do not correspond to the locations of their farm business activity. Several in the survey sampler in fact, received their questionnaires in other states: Arizona, Colorado, Florid” Indiana. Kentuckv, Nevada, and New York.California farms are most commonly organized as &Ole proprietorships. Other forms of organization are, in order of frequency, corporations, family partnerships, and non-family partnerships. In the survey sample, corporations and family partnerships together make up more than half of all farm businesses. All four of these organizational types exist in every farm size group, but farms with larger payrolls are much more likely to be corporations and those with smaller payrolls to be sole proprietorships . Geographically ,the Central Coast region has a larger than average share of corporate farms and non-family partnerships , and a smaller share of sole proprietorships . These farms have been operating uncler their present owners or family predecessors an average of nearly 30 years, one since 1855 and shortest in the coastal regions . Although one respondent operates in more than ten counties, more than four-fifths do all their farming in one county, and more than half of the rest in two . They perform an average of 3.8 farming functions , most frequently harvesting , cultivation and plant care , planting , and land preparation .
Only a third of the businesses market farm commodities. Less than one-fifth limit their operations to a single function, animal products firms by far most likely to be among them. Respondents specified quite a collection of “other” functions as part of their businesses, including manure spreading, catering, beekeeping, selling retail, equestrian training, trucking, financing, holding in cold storage, irrigating, labor contracting, aerial spraying, and almond hulling. Most of these could be — but have not been — interpreted as one of the functions listed on the survey questionnaire .All major crop types are well represented among respondents . Farms that primarily produce tree fruits or other fruits constitute a quarter of those participating. Makmg up about an eighth each are producers of animals, nuts, grapes , vegetables, and “other crops.” Most in the “non-edibles” group indicated that they grow cotton; producers of hay, flowers, silage, and seeds are also in this class. Because their numbers arc quite small, we lumped into the “other crops” group farms that report mainly producing grains, other edible field crops, ornamentals, and other nursery products. Other crops in this group that were specifically indicated by one or more respondents include bees, oysters, herbs, sugar beets, flowers, olives, avocados, horseradish, storks, trout, and meal worms. As with “other” specified business functions,dry racks for weed farms have not been reassigned in the data base from the respondent-indicated crop category to other plausibly appropriate crop types listed on the questionnaire. Relationships between crop type and region are evident. Although most crop types are found in most or all regions, some are concentrated in one or two. Most nu t production is in the central valley regions, grapes in the North Coast and San Joaquin Valley regions , vegetables in the Central and South Coasts, and non-edibles in the San Joaquin Valley and Desert. Agriculture in counties designated as “other” is heavily based on animals, and “other crops” arc an unusually large share of Sacramento Valley farm products.Farms with greatest acreage arc In the Sacramento Valley and counties iksignated “other”, but the Central Coast and Desert tend to have the largest operations with respect to sales value, total labor expense, annual payroll, and peak employment level . Revenues, labor costs, and employment levels arc smallest, on average, in the “other”, North Coast, and Sacramento Valley regions. The ratios of median labor expense to sales value, and median payroll to sales value are rough measures of aggregate labor-intensity of farm production. The ratio of payroll to total labor expense is an inverse indicator of the extent to which labor is obtained from non-employees. On this basis, the findings presented in the table suggest that reliance on labor contractors and custom harvesters is greatest in the South Coast and San Joaquin Valley regions.The work of fann businesses in the survey is performed by an average of 89 workers at peak, 27 year round . Vegetable fanns have by far the largest numbers of workers , and grape and other fruit operations also have much larger work forces than other crop groups. The size difference between seasonal peak and year-round work forces is proportionately greatest in grapes and smallest in animal commodities producticid, There are many roads to farm work, and several types of relationship between fann operator and labor provider.
Two-thirds of all workers in farm businesses, at peak activity as well as year around, are employees on the farm payroll. Farmer family members are only six percent of the overall year-round workforce, two percent at peak, They arc somewhat more likely than not to be on the books as employees, rather than as “unpaid family members.” Larger operators provide the longest periods of employment during a year, over the total survey sample as well as within crop classifications. Significant regression results show that the greater the payroll, the higher the ratio of year-round employees to peak employees, indicating that larger farms tend less to hire and layoff workers around their seasonal variations in need for labor, Employees of farm labor contractors are nearly a quarter of all workers at peak, on average, and one-seventh of workers year-round , FLC employees thus make up almost twice as much of the peak as of the year-round workforce. Although the activity of FLCs is popularly associated with fruit and vegetable production, their employees constitute large segments of the peak work forre in all crop groups except animals. Custom harvester . Only in the tree and other fruit group, where the ratio of employees to all workers on farms was lowest among major commodity classes in 1986, has there been even a slight relative increase in direct farm employment since then. Both the share of farmers in the survey gettmg labor from FLCs and CHs, and the average number of these service providers doing business with each farm reportedly increased from 1986 to 1992. Three in five farm busmcsses purchased services from at least one FLC or CH in 1992. Producers of vegetables were most inclined to use labor contractors, and producers of non-edibles to use custom harvesters . Farmers’ use of licensed pest control operators similarly grew over this period, with two-thirds of vegetable firms and three-quarters of non-edibles firms obtaining service from one or more PCOs.Unwanted turnover raises “arious administrative and supervisory costs, and workforce stability is generally valued. Turnover is both expensive in itself and often a symptom of other problems. Farms experience employee turno’er both during and between production years.