HLFS Summary Tools

The Household labour force survey (HLFS) Summary tools consist of a collection of tools that enable you to construct summaries out of the HLFS survey dataset available on the IDI. The tools are intended for those who would like to construct population-level statistics from the HLFS data, but do not want to wrangle with the survey weights to construct confidence intervals or error estimates for the population statistics.

IMPORTANT NOTE: The outputs produced from this tool will still need to undergo the usual processes of confidentiality/rounding/suppression before you submit these for output checking. The tool does not do any of these tasks for you.

The tools are designed to require minimal understanding of what happens under the hood, and you can use these tools from SAS, R or SQL. The tools require an installation, and detailed instructions on installing these in your project area (& pre-requisites) are available in the HLFS Data Tools User Guide. The tools are written in SQL (and must be installed on the SQL server), but you do not need SQL expertise to install or run these.

Each tool allows you to construct a very specific kind of output summary. There are 6 such tools currently available as part of this package:

• HLFS Levels Summary Tool (SELevel): This tool constructs weighted population counts grouped by a set of user-specified variables, along with the confidence intervals around the weighted counts.

• HLFS Level Changes Summary Tool (SEPercentChange): This tool generates the difference in weighted population counts from two subsets of the HLFS data (for instance, two different HLFS quarters), along with confidence intervals around this difference. There is also an input parameter that enables you to output difference & confidence intervals expressed as percentages, if so required.

• HLFS Means Summary Tool (SEMean): This tool constructs weighted means for a numeric measure (like earnings or hours worked) grouped by a set of user-specified variables, along with the confidence intervals.

• HLFS Mean Changes Summary Tool (SEMeanChange): This tool generates the difference in weighted mean for a numeric measure, grouped by a set of user-specified variables across 2 slices of the HLFS data (say, 2 different HLFS quarters). It also provides the confidence intervals around this difference estimate.

• HLFS Rates Summary Tool (SERate): This tool generates the prevalence rates of a variable as measured against a denominator grouped by a set of user-specified variables, along with the confidence intervals around the rate. For instance, one can measure the “not-employed” rate expressed as a percentage of all other labour force status values.

• HLFS Rate Changes Summary Tool (SERateChange): This tool generates the difference between prevalence rates of a variable as measured against a denominator, from 2 different subsets of the HLFS data (say, 2 different HLFS quarters), grouped by a set of user-specified variables. It also provides the confidence intervals around this difference estimate.

Usage, Input and Output Specifications

The tools require the user to supply a set of inputs - like the HLFS dataset name, the filters to be applied, the grouping columns, etc. and it returns a table with the summary estimates, confidence interval and relative sampling errors. The exact input & output specifications are slightly different for each tool, and can be obtained from the tool-specific page linked above. The pages also provide examples of how to use the tool from SQL, SAS and R.