A goal of Wyoming's Labor Market Information office has been to develop a low cost state specific alternative to national surveys such as the Current Population Survey (CPS), and various longitudinal surveys. In addition to their high cost, such national surveys lack sufficient sample to draw inferences about state and local population characteristics and labor market dynamics. Moreover, national surveys (e.g., the CPS) sample physical locations rather than the state's population and face difficulties tracking geographically mobile sample members. State developed administrative data bases provide a mechanism circumventing many of these problems. With ES-202, Wage Records, and other files we have created rich and comprehensive databases suitable for the most sophisticated analysis of labor economics, employment and training program evaluation, and the related needs of Workforce Investment Boards and One Stop environments.
The administrative records approach documented in this publication represents over five years of effort. The publication includes many new concepts, and contains basic information about the organization and management of data elements. However, it omits examples from ongoing research into the dynamics of employment hires and retention, and the linkage of these time series to current employment estimates and projections. As a baseline document, Wyoming Wage Records is intended to describe basic file content and the development of key concepts and data sets without which analysis would be impossible.
Administrative records represent a mass of data. This publication chronicles the work of many in organizing these data elements into meaningful classification systems whose test as useful concepts is yet to be realized. In contrast to obtrusive measures, such as the CPS, administrative records contain none of the sampling and non-sampling error associated with surveys. Detecting systematic bias and error in these administrative data rests, almost exclusively, on work involving the development of classification systems and basic tests of consistency among merged data sets. Developing the classification system for multiple job holders creates both the norm against which error can be detected in nonconforming data, and a variable which must have face, concurrent, and/or construct validity in cross tabulation analysis with industry, earnings, and basic demographics. Replication of these tests, we hope, will soon take place in other states.
Finally, the underlying theoretical emphasis in the development of administrative data has been pragmatic. We view workers and employers as goal directed and engaged in strategies based on learned behavior. Patterns of work and hiring are associated with market position measured in terms of seller demographics, purchaser characteristics such as industry niche and size, and the ecology of the labor market. One of the most difficult tasks before us is localizing supply and demand functions so that the relevant market ecology can be described. The second task is to tie market dynamics (e.g., hire and retention rates) to current employment estimates and projections. Acquiring and using state administrative data allow us to break away from what now seems archaic ways of viewing the labor market, and the limitations of national sample surveys with sponsors having interests far from state and local needs.
|Go Back To:|
|Table of Contents|
|Labor Market Information|
|Wyoming Job Network|
|Send Us Mail|