Skip to content

Update on the Current Population Survey Research

Mon Aug 01 2016
Stephanie Chan-Yang, Yang Cheng and Aaron Gilary, U.S. Census Bureau
Component ID: #ti802791762

The U.S. Census Bureau’s Current Population Survey is one of the oldest and largest household surveys in the United States. Since 1940, it has produced monthly statistics on labor force information. The Current Population Survey interviews about 72,000 households each month to estimate totals of persons unemployed, employed and not in the labor force, leading to the official estimate of the national unemployment rate.

Component ID: #ti1544601902

The Current Population Survey applies a stratified two-stage cluster sampling design to select a representative sample of U.S. households. A housing unit selected for the sample is interviewed for four consecutive months, rotated out for eight months, and then interviewed for another four months. This approach aims to develop overall monthly estimates while also tracking monthly and annual changes among the sampled households.

Component ID: #ti1696253601

To manage these design features, the survey team has long relied on cutting-edge research on sampling, weighting and variance estimation. Given several key variables that are beyond our control such as budgets, computational power and stakeholders’ needs, the survey team must be sufficiently agile to adapt to changes. This nimble approach requires a strong understanding of the underlying theory, an ability to adapt the survey quickly, and an opportunity to hone our methods under peer review.

Component ID: #ti1696253600

Research Presented at the Joint Statistical Meetings

Component ID: #ti1696253599

The survey team will give three presentations in the session “Update on the Current Population Survey Research” at this year’s Joint Statistical Meetings in August 2016.

Component ID: #ti1696253598

  • Stephanie Chan-Yang will speak about the sample size for the Current Population Survey. She will also explain the sample size and allocation in relation to the Bureau of Labor Statistics sample design requirements for accuracy. Chan-Yang will further describe the Children’s Health Insurance Program expansion to the survey sample. This expansion increases the survey sample size in order to provide better estimates of low-income children without health insurance. These data feed into the Current Population Survey Annual Social and Economic Supplement. Finally, Chan-Yang’s presentation will explore recent research on reducing the sample size and budget constraints.
  • Yang Cheng will explore a new method to improve our composite estimates. In his research, he proposed an iterative version of our composite estimator (known as the AK composite estimator) for the Current Population Survey. This new method includes the current AK composite estimator as a special case. In addition, the proposed method will reduce the mean squared error of the AK composite estimator when we choose the optimal estimator in this general family. Finally, Cheng will demonstrate the proposed method via comprehensive numerical studies.
  • Aaron Gilary will give an overview of Current Population Survey variance methodology. This talk discusses current survey methods of calculating variances, with a focus on the Balanced Repeated Replication method. This method is used to construct a variance estimate by resampling the data using replicate factors. The talk highlights the components of the variance estimate that come from the survey sample design, and the different variance measures that the survey produces. We will conclude the presentation with ideas for improvement in the future.

Component ID: #ti1696253597

To learn more about Current Population Survey methodologies and researches, please join us at the Joint Statistical Meetings on August 3, 2016, or contact us at: <stephanie.chan.yang@census.gov>, <yang.cheng@census.gov>, or <aaron.j.gilary@census.gov>.

X
  Is this page helpful?
Thumbs Up Image Yes    Thumbs Down Image No
X
No, thanks
255 characters remaining
X
Thank you for your feedback.
Comments or suggestions?