Research Discussion Paper – RDP 2021-08 Job Loss, Subjective Expectations and Household Spending

This ‘read me’ file contains information on the files used to generate the results presented in RDP 2021-08.

Our analysis was done using the statistical programming language Stata. We used version 16.0. In order to help users to replicate our work, we have included the Stata project file ‘rdp-2021-08-supplementary-information.stpr’. Users can use this file to replicate our work as follows: Open Stata, select File-Open and then select the ‘rdp-2021-08-supplementary-information.stpr’ file.

Our codes are distributed as is, without warranty, and are solely for the purpose of replicating our results. Any alternative use of the code is not supported.

Data

• HILDA

  The majority of the analysis presented in the paper comes from the Household, Income and Labour Dynamics in Australia (HILDA) Survey. The HILDA Survey provides individual-longitudinal level data. Data from the HILDA Survey is available to researchers living in Australia or overseas. Data may only be accessed through the National Centre for Longitudinal Data Dataverse. For more information on data access, please see <https://melbourneinstitute.unimelb.edu.au/hilda/for-data-users>.

• ABS

  Data in Figure 1 are constructed using publicly available data from the Australian Bureau of Statistics (ABS) release ‘Labour Force, Australia, Detailed’, and are derived from employment, and employment by expectations of future unemployment (not to be with current employer in 12 months due to: employer closing down or downsizing; seasonal, temporary, fixed contract or casual; and finish work for other reasons).

The file ‘rdp-2021-08-graph-data.xlsx’ contains the publically available data used to directly plot the figures.

Stata project file

‘rdp-2021-08-supplementary-information.zip’ does not include data to run the code, as this must be requested and obtained from the National Centre for Longitudinal Data Dataverse. Once data has been acquired, saving the file ‘Stata’ as received from the NCLDA in the same file as the working directory of the replication file will allow the below code to run.

Code

The Stata project file collates a series of do files. To produce the necessary output, do files should be run in the following order:

• Merge data

  Do file: 1-merge.do

  Functions: Creates an unbalanced longitudinal data file using the raw combined files. The resulting data file is saved in Stata's long format. Uses ‘tempfile tempdata_w’ to create a local macro that allows us to access a temporary data file which will be automatically deleted when the do file ends.

  Input: Files Combined_a190u.dta, Conbined_b190u.dta through Combined_s190u.dta. These originate from the ‘Stata’ file provided by the NCLADA.

  Output: long-file-unbalanced.dta

• Clean data

  Do file: 2-clean-data.do

  Functions: Cleans data and creates variables for use in later analysis.

  Input: long-file-unbalanced.dta

  Output: long-file-cleaned.dta

• Analysis 1

  Do file: 3-analysis-1.do

  Functions: Creates all figures and regression output in Sections 1, 2 and 3 of the RDP, which covers subjective expectations.

  Input: long-file-cleaned.dta

  Output: figure1.xls, figure2.xls, figure3.xls, figure4.xls, figure5.xls, figure6.xls, figure7.xls, figure8.xml

• Analysis 2

  Do file: 4-analysis-2.do

  Functions: Creates all figures and regression output in Sections 4 and 5 of the RDP, which cover the consumption response to unemployment.

  Input: long-file-cleaned.dta

  Output: figure9.xls, figure10.xls, figure11.xls, figure12.xls, table2.xls, figure13.xls, figure14.xls, figure15.xls, figure16.xls, figure17.xls, figure18a.xls, figure18b.xls

• Analysis appendices

  Do file: 5-analysis-appendix.do

  Functions: Produces all figures and regression output for the Appendices.

  Input: long-file-cleaned.dta

  Output: figurea1.xls, figurea1.dta, figureb1.xls, figureb1.dta, tableb1.xml, tableb1.txt

18 August 2021

Back to abstract