The Office for National Statistics Longitudinal Study

Main Article Content

Alison Sizer
Oliver Duke-Williams

Abstract

Background and Rationale
The ONS Longitudinal Study (‘the LS’) covers England and Wales and includes individual data from the 1971 – 2011 decennial censuses and linked information on births, deaths and cancer registrations. It is representative of the population of England and Wales.


Aim
This presentation describes the LS and the linked administrative data, and showcases recent/ prominent examples of research.


Methods and Approach
The LS is built around samples drawn from decennial censuses, with its initial sample drawn from the 1971 Census. It also contains information about other people living in a sample-member’s household. Substantial emphasis is placed on security of access to the data and its responsible use. All research outputs are checked and are only released to users once disclosure control requirements are met. Linkage of study members from one census to another and vital events is carried out by ONS.


Results
The LS has been used for a variety of research. Using linked census and death records occupational differences in mortality rates have been researched. Individual records from all five censuses have been used to contribute to research social mobility, and research has also investigated the effects of long-term exposure to air pollution. Research has provided evidence of impact for social policy issues, e.g. health inequalities and the State Pension Age Review.


Discussion
The main strength of the LS is its large sample size (>1 million), making it the largest nationally representative longitudinal dataset in the UK. This allows analysis of small areas and specific population groups. Sampling bias is almost nil, and response rates are very high relative to other cohort and panel studies.


Conclusion
The ONS Longitudinal Study is a vital UK research asset, providing access to a large sample of census data linked across five censuses. It is strengthened through linkage to events data.

Background and Rationale

The ONS Longitudinal Study (‘the LS’) covers England and Wales and includes individual data from the 1971 – 2011 decennial censuses and linked information on births, deaths and cancer registrations. It is representative of the population of England and Wales.

Aim

This presentation describes the LS and the linked administrative data, and showcases recent/ prominent examples of research.

Methods and Approach

The LS is built around samples drawn from decennial censuses, with its initial sample drawn from the 1971 Census. It also contains information about other people living in a sample-member’s household. Substantial emphasis is placed on security of access to the data and its responsible use. All research outputs are checked and are only released to users once disclosure control requirements are met. Linkage of study members from one census to another and vital events is carried out by ONS.

Results

The LS has been used for a variety of research. Using linked census and death records occupational differences in mortality rates have been researched. Individual records from all five censuses have been used to contribute to research social mobility, and research has also investigated the effects of long-term exposure to air pollution. Research has provided evidence of impact for social policy issues, e.g. health inequalities and the State Pension Age Review.

Discussion

The main strength of the LS is its large sample size (>1 million), making it the largest nationally representative longitudinal dataset in the UK. This allows analysis of small areas and specific population groups. Sampling bias is almost nil, and response rates are very high relative to other cohort and panel studies.

Conclusion

The ONS Longitudinal Study is a vital UK research asset, providing access to a large sample of census data linked across five censuses. It is strengthened through linkage to events data.

Article Details

How to Cite
Sizer, A. and Duke-Williams, O. (2019) “The Office for National Statistics Longitudinal Study”, International Journal of Population Data Science, 4(3). doi: 10.23889/ijpds.v4i3.1205.