This booklet contains tabulated results of the 2006 Wage and Benefit Survey conducted by the Farm Employers Labor Service (FELSŪ) and eight other organizations in collaboration with UC Cooperative Extension. We appreciate the participation of respondents and managers from the following associations:

Thanks also to Brenda Trammel (California Farm Bureau Federation) for data entry and Gary Casterline (UCB College of Natural Resources) for database input and report programming.

As an aid to interpreting the survey reports, the explanation below describes terms used and aspects of data treatment.

Wages Section:

The tables show the overall average, the average low and high, and the absolute low and high wages reported for each job by respondents in the specified associationXcommodity group. They indicate not only the total number (N) of respondents reporting any wage for each respective job-paytype (pay for eight job titles could be reported per hour, month, or both) but also the number (n) who report paying a single rate (lowest=highest) for each job. The standard deviation (sd) is of the midpoint between the lowest and highest wage rate for the job reported by each respondent. The smaller the standard deviation, the closer that wages paid by all responding firms for the given job tend to cluster around their average.

Each Job Title was briefly defined on the survey form (page 2) so that survey participants could report wages relative to similar job content.

Benefits Section:

Changes in the survey form after 2003 affect the meaning and year-to-year comparability of results shown in the benefits section. While check boxes for both "yes" and "no" responses were included in surveys through 2003, respondents could only indicate "yes" in 2004 and 2005, and the standard for an affirmative response was clarified to "received by more than half of employees without their having to co-pay more than 25% of the cost." The portion of replies with a "yes" for each benefit was calculated before 2004 considering a base (denominator) of only those who checked either of the two boxes, i.e., ignoring those who left the item blank. Because the portion is now calculated in relation to all respondents, including both those who would have left the item blank and those who would have explicitly checked "no," the 2004-2005 figures tend to be lower than pre-2004. N is the total number of respondents in the specified associationXcommodity tabulation group.

Workers' Compensation Insurance Premium Expense Section:

The reports show the percentage change [current year premium minus previous premium] divided by [previous premium], considering only respondents who report both previous and current year premiums. N is the number of such respondents. There is no control for year-to-year differences in number of employees covered.

We hope you find these survey results helpful. Any suggestions to improve this service are welcomed.