Surveys of businesses and other organizations are difficult to develop and implement. The challenges faced when conducting surveys of businesses are very different from those encountered in conducting other types of surveys. In this section we briefly discuss three key challenges that should be considered when developing the surveys of employers and labor market intermediaries: sample design, survey administration, and employer-intermediary linkages.
The first and most obvious challenge in conducting surveys with employers is identifying an appropriate sample frame, or list, from which to draw the sample. Ideally, the sample frame should include all or nearly all members of the population of interest in this case all domestic organizations with employees.(22) A survey could be limited to employers in the private sector or could include public organizations as well. Holzer's survey of employers included both groups.
Assembling a complete list of businesses is challenging. Non-government lists are generally constructed from a range of directories (e.g., phone books) and files (e.g., lists of trade organizations); government lists (such as those maintained by the U.S. Census Bureau and the Bureau of Labor Statistics) are generally derived from tax, unemployment insurance, or other records. Regardless of the source, it is difficult to construct and maintain an up-to-date list of all U.S. employers, and most existing lists suffer from a range of defects such as duplicate records, classification inaccuracies, incompleteness, and extraneous units (Srinath, 1987).
Developing a sample of businesses is further complicated by the difficult task of identifying the right organizational entity to which the survey should be directed. Business organizations include enterprises that consist of one or more companies, companies that consist of one or more establishments, and establishments. Employer surveys, such as Holzer's Multi-City Telephone Employer Survey (MCTES) and later four-city survey, have sampled establishments, recognizing the local nature of many businesses' employment practices. However, determining whether given business organizations are establishments, rather than enterprises or companies, can be difficult. The sample frame must be systematically "profiled" to identify and code multi-unit business organizations so that establishments may be sampled (Srinath, 1987).
5.1.2 Survey Administration Challenges
Researchers typically encounter additional difficulties in their attempts to survey businesses. The difficulties include the following.
Respondent Identification. Identifying the best person(s) to complete a survey within a business organization is challenging. The information obtained from the lists typically used for sampling generally includes the company's address and telephone number and, in some cases, the name of a general contact person in the organization. Unfortunately, the identified contact is rarely the person to whom the survey should be directed that is, the person who knows the most about entry-level worker recruitment and hiring. Moreover, elements of the survey may need to be administered to more than one person within the business organization. For example, it may be most appropriate for a person who handles recruitment to answer one survey module and another person responsible for employment and supervision to respond to a second module.
Gatekeepers and Corporate Survey Policies. In many businesses there are "gatekeepers" who screen mail and telephone requests for survey participation and prohibit access to the survey's intended respondent. In addition, organizations often have "corporate survey policies" that dictate whether and in what ways an employee may respond to a survey. Generally speaking, these policies apply to certain questionnaire methods (such as written questionnaires versus in-person interviews) and require potential respondents to obtain permission to participate in the survey from a central authority (Dillman, 2000). However, some companies have "no survey" policies that prohibit employees from participating in surveys on the company's behalf.
Incentives for Survey Participation. Surveys of business organizations often struggle to obtain meaningful response rates. Employees often are busy and may have few personal incentives to respond to a survey about employer practices (Dillman, 2000).
5.1.3 Linking Employer Practices with Intermediaries
As discussed in previous chapters, many employers utilize labor market intermediaries in their recruitment, hiring, and employment processes. The simplest way to obtain information about the use of intermediaries is to rely entirely on information supplied either by employers in an employer survey or by intermediary organizations in a survey of labor market intermediaries. Holzer's four-city surveys in the late 1990s asked employers about the involvement of intermediaries in their employment activities. Abt Associates' surveys of intermediaries during the same period asked these organizations about the employers with whom they worked. However, neither set of surveys asked detailed questions about the interactions between employers and intermediaries.
A more demanding, but potentially very rewarding, approach would be to ask both employers and intermediaries about their interactions. Specifically, a national survey might ask employers about the intermediaries with which they have worked, and then interview the intermediaries named by employers. By doing this, the survey would provide more thorough information than is now available. This approach also would allow verification of the survey responses of both employers and intermediaries.
"report.pdf" (pdf, 212.99Kb)