Contents

---

Introduction

Background

Current Design of the SESTAT Surveys

Alternative Approaches

Summary

Appendix A. Workshop Agenda

Appendix B. List of Participants

Appendix C. Expert Panel Meeting Summary

Appendix D. Desired and Acceptable CVs for Estimates of the S&E Population

Appendix E. Characteristics of Establishment-Based Data Collections

Bibliography

Suggested Citation, Acknowledgments

Introduction

Knowledge about the size and characteristics of the science and engineering (S&E) workforce is extremely important in understanding and assessing the place of the United States in this increasingly technological world. For more than 50 years, the National Science Foundation (NSF) has attempted to meet the need for current information on this important segment of our workforce. Although representing only about 4.6% of the civilian noninstitutionalized population in 1997, or about 12.5 million people, the impact of this group on our society far exceeds its number because of the contributions that scientists and engineers make to technological innovation and economic growth. It is no exaggeration to anticipate that scientists and engineers will prove critical to the continued social and economic development of the nation as it confronts the challenges of the future.

The most comprehensive current source of data on the S&E workforce in the United States is the integrated Scientists and Engineers Statistical Data System (SESTAT), a revised system that NSF introduced in 1993. SESTAT draws upon three surveys, with data collection carried out in 1993 and then at 2-year intervals for the rest of the decade. The three surveys are as follows:

• National Survey of College Graduates (NSCG). In 1993, the NSCG sample was drawn from all people age 75 or younger who had a bachelor's or higher degree at the time of the 1990 decennial census. Certain occupations and groups were oversampled because a main purpose of the survey was to screen for individuals age 75 or younger with a bachelor's degree or higher in an S&E field at the time of the 1990 decennial census (whether employed in S&E or not in 1993), and those with a bachelor's degree or higher in non-S&E disciplines in 1990 who were employed in an S&E occupation in 1993. These two groups comprise what is termed the NSCG S&E panel that was followed in 1995 and later rounds.
  
  
• National Survey of Recent College Graduates (NSRCG). The NSRCG covers those individuals who received an S&E bachelor's or master's degree from a U.S. educational institution within the academic years 1 or 2 years prior to the survey reference date. A sample from each cohort of the NSRCG is resurveyed every 2 years thereafter, using the same questionnaire as the NSCG.
  
  
• Survey of Doctorate Recipients (SDR). The SDR covers individuals age 75 years or younger who received a doctorate degree in an S&E field from a U.S. educational institution since June 1942. The SDR sample of S&E doctorate recipients is subsampled over time, with samples of new S&E doctorate recipients added every 2 years. The subsampling rates for earlier doctorate recipients and the sampling rates for recent doctorate recipients are chosen to maintain a sufficient total sample for analysis purposes. The panel design produces longitudinal data that may be used to analyze the professional careers of S&E doctorate recipients.

Thus, the SESTAT target population includes residents of the United States who were noninstitutionalized and age 75 or younger with at least a bachelor's degree who were either educated as or working as a scientist or engineer.

Although SESTAT provides substantial employment-related data for those in the target population, users note some limitations in the data. The current definition of the SESTAT target population is restricted to those with at least a bachelor's degree. Thus, individuals with a science or engineering occupation but who do not have a bachelor's or higher degree (in any field) are excluded from the target population.

In addition, the current NSF system has some problems covering the S&E target population. Although those residing in the United States at the time of the 1990 decennial census with only non-U.S. degrees are covered by SESTAT through the NSCG, those with degrees who currently reside in the United States but came here after the census, and who did not receive a degree from a U.S. institution since 1990, are not covered. Also, individuals holding non-S&E degrees in 1990 and working in non-S&E occupations in 1993 who changed to S&E occupations after that time are not covered. Furthermore, individuals working in an S&E occupation who received non-S&E degrees since 1990 are not covered.

This report addresses issues of target population and coverage and examines the feasibility and desirability of other survey approaches for obtaining the required data with the desired sample sizes. Other existing federal statistical data collection efforts are considered for use as possible sources of samples for the SESTAT surveys. Thus, the report considers alternative methods for addressing two issues: (1) defining the target population to meet user data needs, and (2) achieving good coverage of the target population.

As noted above, being a small proportion of the total population, scientists and engineers are a "rare" population, i.e., they are difficult to study in a cost-effective manner using general population survey techniques. For example, to locate individuals in the general population who are in scope for the S&E population, and to oversample particular subgroups of the S&E population, would require a large volume of household screenings. It is essential, therefore, to find a cost-effective approach to screen the general population. In this context, it is natural to look at a variety of alternative sample designs, as well as to examine existing large-scale, ongoing governmental household surveys that might lend themselves effectively to such a screening requirement. To that end, this report explores and examines the feasibility of using several federal data-gathering efforts for screening purposes, including the planned American Community Survey, the Current Population Survey, the National Health Interview Survey, the National Immunization Survey, and other selected large household or population surveys. Another very different approach is to sample members of the S&E population at their places of work, using an establishment-based sample design. This report also explores this option as a source of data on the S&E workforce.