nsf.gov - NCSES Comparison of the National Science Foundation's Scientists and Engineers Statistical Data System (SESTAT) with the Bureau of Labor Statistics' Current Population Survey (CPS) - US National Science Foundation (NSF)
text-only page produced automatically by LIFT Text
Transcoder Skip all navigation and go to page contentSkip top navigation and go to directorate navigationSkip top navigation and go to page navigation
National Science Foundation National Center for Science and Engineering Statistics
Comparison of the National Science Foundation's Scientists and Engineers Statistical Data System (SESTAT) with the Bureau of Labor Statistics' Current Population Survey (CPS)

Summary and Conclusions

 

The two main purposes of this report are (1) to compare the SESTAT and CPS estimates of the S&E workforce with bachelor's or higher degrees and (2) to obtain estimates from CPS of the numbers of individuals in S&E occupations who do not have a bachelor's or higher degree. For the first purpose, comparisons between SESTAT and CPS were restricted to the subset of the S&E population with bachelor's or higher degrees who are working in S&E occupations. The remainder of the SESTAT population, people with S&E degrees who are not working in S&E occupations, cannot be identified using CPS data because CPS does not collect the data on degree fields needed to identify S&E degrees. Therefore, no comparisons between SESTAT and CPS could be made for this group. CPS was used for the second purpose because individuals without a bachelor's or higher degree are not included in SESTAT. The principal findings and conclusions of this study are summarized below.

  1. SESTAT is a longitudinal data system that follows a large sample of individuals (about 100,000 in the 1997 cycle) with S&E bachelor's or higher degrees or S&E occupations over time. It is a rich data system for analyses of these individuals. CPS is a major survey that provides up-to-date information on the labor force and demographic characteristics of the general, civilian, noninstitutionalized U.S. population. SESTAT includes individuals with a bachelor's or higher degree and CPS includes these individuals along with those who have a degree below a bachelor's. This report points out differences in the two data sources and explores the possible use of the CPS data to provide limited information about individuals below a bachelor's degree who are employed in S&E occupations.
  2. Differences exist in the mode in which SESTAT and CPS data are collected, which may affect nonsampling errors. In the 1997 SESTAT, 61% of responses were obtained by mail, 37% by CATI, and 2% by CAPI. CPS uses a combination of CAPI and CATI collection; CAPI is used during the first and fifth months of participation and CATI is used in the other months of participation. SESTAT and CPS also differ in the use of proxy respondents; all SESTAT sample members are asked to provide self-reports, whereas CPS relies heavily on proxy reports. However, it is unlikely that there is an appreciable mode effect for the items considered in this analysis.
  3. Differences exist in the way information on educational attainment, occupation, and certain demographic data are collected and coded in the respective survey systems. Some of these differences are summarized as follows:
    • In the collection of data on degrees, the main difference between CPS and SESTAT is that CPS collects the highest level of school or degree completed and does not collect the field of degree, whereas SESTAT collects the school, degree level, date, and field for degrees at the bachelor's or higher level. SESTAT also collects some information on associate's degrees.
    • The occupation-coding process is different for the two survey systems. In CPS initial interviews, coders assign occupation codes using verbatim descriptions of the occupation, job duties, and industry. During follow-up months, dependent interviewing is used in CPS and data collected and coded during previous months may be used without coder review. In SESTAT, respondents provide verbatim responses and then select their own occupation codes, either from a printed list or through a series of CATI screens. Coders review these verbatims and self-selected codes along with other survey data, but they change the respondent's code only when sufficient information exists to allow the assignment of a better code. Different occupational taxonomies are used in the two survey systems, but they were derived from the same source and are generally consistent with each other.
    • Both SESTAT and CPS collect data on race in a similar manner, but the collection of data on ethnicity is different. The three SESTAT sources of Hispanic-origin data (1990 census, SED, NSRCG) ask directly whether the respondent is of Hispanic origin (with slightly different wording). In CPS, respondents are asked to select their origin (or the origin of some other household member) from a "flash card" listing 20 ethnic origins. Individuals of Hispanic origin are those who indicate their origin is Mexican American, Chicano, Mexican (Mexicano), Puerto Rican, Cuban, Central or South American, or other Hispanic. That is, Hispanic origin is self-reported in SESTAT and coded from origin data in CPS.
  4. The total S&E population in the 1997 SESTAT includes more than 12 million individuals who (1) have a bachelor's or higher degree in an S&E field or (2) have a bachelor's or higher degree in a non-S&E field and had worked in an S&E occupation in the 1993 NSCG. The definition of the S&E population in SESTAT depends on both degree field(s) and occupation. CPS, which is based on a nationally representative sample of the civilian, noninstitutionalized population of the United States, provides an alternative source for studying the S&E population. However, CPS does not collect data on degree field. Thus, although CPS conceptually covers the same S&E population as SESTAT (with the exceptions noted below), it can only be used to study a subset of the S&E population, namely, the roughly 3 million individuals with bachelor's or higher degrees who are employed in S&E occupations. CPS can also be used to obtain estimates for the population working in S&E occupations without a bachelor's or higher degree.
  5. The coverage of the S&E population in SESTAT and CPS differs in several ways:
    • SESTAT does not cover a portion of "foreign-trained" scientists and engineers, that is, U.S. residents who (1) were not eligible for inclusion in the 1993 NSCG who received an S&E bachelor's or higher degree from a non-U.S. institution only or (2) hold a non-S&E bachelor's or higher degree from a foreign institution and entered the United States after the 1990 census and later obtained an S&E job.
    • SESTAT excludes individuals currently working in S&E occupations who do not have a bachelor's or higher degree in an S&E field unless they were working in an S&E occupation in 1993.
    • With the exception of the March supplement, CPS excludes all active-duty military personnel. SESTAT, on the other hand, attempts to collect data from sample individuals in the military if they are living in the United States during the reference week.
  6. The April 1997 CPS indicates that about 210,200 (6%) of the estimated 3.5 million individuals working in S&E occupations entered the United States during or after 1990 (table 5Excel table.).[12] This estimate includes individuals who may have earned an S&E bachelor's or higher degree from a U.S. institution after April 1990 (and thus would be included in the NSRCG and/or SDR components of SESTAT). It also includes those who entered the United States between January and April 1990 and thus would be covered in NSCG. Therefore, the estimated 210,200 recent immigrants (SE = 25,000) provide a rough upper boundary on the number of immigrants from 1990 or later who are excluded from SESTAT.
  7. The SESTAT data system indicates that an estimated 594,000 individuals with non-S&E bachelor's or higher degrees were employed in S&E occupations in 1993 (table 1Excel table.). The SESTAT estimate for this group decreased to 334,100 in 1995, and then to 292,000 in 1997. The drop-off reflects the fact that the SESTAT design does not include provisions for updating the original sample of those in S&E occupations who do not have a bachelor's or higher degree in an S&E field. Assuming that the actual number of such individuals has not decreased since 1993, an estimated 302,000 or more people may not have been covered in the 1997 SESTAT data system.
  8. After accounting for the coverage differences noted earlier to the extent feasible, weighted estimates of the numbers of individuals with a bachelor's or higher degree who are employed in S&E occupations derived from the two data sources are roughly comparable for 1997: the SESTAT estimate is 3.35 million (SE = 26,600) and the CPS estimate is 3.33 million (SE= 98,700; table 9Excel table.).
  9. Individuals who do not have a bachelor's or higher degree who are working in S&E occupations are not included in SESTAT by definition. In the April 1997 CPS sample, only 431 individuals without 4-year college degrees were working in S&E occupations (appendix C, table C-18). On a weighted basis, these 431 sample individuals represent about 997,600 individuals without bachelor's degrees who are working in S&E occupations (SE= 54,400; table 12Excel table.). The majority of these people were employed in computer/mathematics or engineering occupations.
  10. The cumulative response rates in SESTAT are generally lower than the corresponding response rates in CPS. Thus, the SESTAT estimates may be subject to somewhat greater nonresponse biases than the CPS estimates. Because portions of SESTAT are panels that are followed over time, response rates in each SESTAT panel may decrease in each subsequent round of data collection. Although the weighting adjustments used in both survey systems are designed to minimize nonresponse biases, complete elimination of nonresponse biases is not certain and comparisons of estimates between the two survey systems may be affected. On the other hand, the large amount of information available for nonrespondents who were surveyed in the past is used to improve nonresponse adjustment techniques in SESTAT.
  11. The sample sizes for CPS are considerably smaller than those for SESTAT for the population of interest. For example, in the April 1997 CPS sample, there are 1,541 individuals with bachelor's or higher degrees in S&E occupations, compared to 48,488 in the 1997 SESTAT (table 16Excel table.). Thus, the CPS estimates are subject to much larger sampling errors than the corresponding SESTAT estimates. Even if several CPS samples are accumulated over time, the effective sample size will not increase appreciably because of the considerable overlap between successive monthly samples. The relatively small sample size of CPS limits the detail of analyses of the S&E population. However, CPS can provide useful information about individuals who do not have a bachelor's or higher degree, such as the number who are working in S&E occupations by occupation category. For example, combining the CPS information about individuals without a bachelor's or higher degree with an analysis of the SESTAT population for those with at least a bachelor's degree may provide limited information for the study of S&E graduates as they make their transition into S&E occupations.

Top of page. Back to Top



Footnotes

[12] Numbers are rounded to hundreds.


 
Comparison of the National Science Foundation's Scientists and Engineers Statistical Data System (SESTAT) with the Bureau of Labor Statistics' Current Population Survey (CPS)
Working Paper | SRS 07-205 | August 2007