Complex Sample Surveys: Design, Estimation, Dissemination

As society increasingly prioritizes equity, inequality of opportunity (IOP) has gained significance. IOP arises when desirable outcomes are unequally distributed due to circumstances beyond an individual's control rather than personal effort. Existing measures of absolute and relative IOP often assume simple random sampling and rely on inequality metrics that ignore survey structure. This study proposes using the Gini index to estimate relative IOP under a complex household survey framework. Additionally, we introduce sequential sampling, an economical approach that optimizes sample size while minimizing estimation error. We analyze the asymptotic properties of this strategy without assuming a specific population distribution, construct confidence intervals for relative IOP, and illustrate our findings using data from India's National Sample Survey Organization (NSSO). 

The Job Opening and Labor Turnover Survey (JOLTS) program of the Bureau of Labor Statistics (BLS) has since 2021 begun to produce state-level estimates to satisfy data users' interest. While JOLTS state-level estimates have proven to be very useful and well received, there is great interest in smaller areas, especially state-industry and sub-state.

We will discuss the feasibility of producing JOLTS small area estimates. JOLTS state estimates use a Quarterly Census of Employment and Wages (QCEW) based model that imputes JOLTS microdata for all QCEW records based on employment change, establishment size and industry classification. The state model estimates are summed within each US Census Region. The percentage share of each state of the regionally summed estimates are applied to the published JOLTS regional estimate to derive a final model state estimate. Essentially, the JOLTS published regional estimates are apportioned to each state based on their model percentage share. Small area estimates could be modeled and apportioned in much the same way, apportioning the JOLTS state estimate to each substate small area based on their QCEW model percentage share. 

Official statistics is often used by governments for policy review and development. It can also be used by the general public for decision-making. However, statistics can sometimes appear too complex for the layman to comprehend.
The Singapore Ministry of Manpower aims to make labour statistics more intuitive and accessible to the masses. Firstly, we conduct career talks and information sessions at tertiary institutions to share information about the job market and how students can use statistics to facilitate their job search. Secondly, we organize statistical briefings to journalists to explain the context behind each statistical release and to clear common misconceptions. Thirdly, we incorporate interactive tools in our statistical website so that users can customize the type of information they are looking for. Lastly, we leverage social media platforms to produce bite-sized content to appeal to people from all walks of life.
In the paper, we seek to bridge the gaps that exist between official statistics and the population. Through a suite of initiatives, we are able to provide statistics that can aid in decision-making and improve statistical literacy within the population.