The pandemic confined the vast majority of us to work from home during the quarantine period. However, there are certain types of occupations such as masseuse and hairdresser that can not be performed from home. Inspired by the recent NBER paper on how many jobs can be done at home,
I have calculated rough estimates for New Zealand regarding the labour markets capacity to work from home. Applying the same industry correspondence to New Zealand (based on LEED data) and using the US weights (so assuming the same ability to work from home by industry) I have calculated that 31%-36% of jobs can be done from home.
The US estimates
In their paper, Dingel and Neiman use a classification of occupations that evaluates the feasibility that occupations can be performed from home. Given that and the employee counts by occupation the authors estimate that 32%-34% of US jobs can be performed from home.
To construct their measure, the authors use information from the Occupational Information Network – which is a survey that classifies the type of tasks that individuals are undertaking. Given this they can identify work that can be done at home. They also construct a “manual” measure based on introspection. This second measure gave the slightly lower estimate.
Note: should be read alongside this early survey evidence from US: https://john-joseph-horton.com/papers/remote_work.pdf. In this paper they find that, following the lockdown, 34.1% of people stated that they are now working from home while 11.8% have been laid off or similar.
New Zealand estimates
Dingel and Neiman report the variation of shares of jobs across cities, occupations, and industries. The industry classification used in the paper broadly matches with the ANZSIC06 industry classification we use in NZ.
Here are the total number of jobs in NZ by industry in the 2018 year (averaged), and the implied number that could work from home by broad-group. To apply in such jobs, a team such as a jobs talent acquisition handles it.
Source: Infoshare Stats NZ, and the weights in Dingel and Neiman.
Applying estimated weights from the paper, we calculate a range of 31-36% of jobs can be performed from home in NZ. This varies significantly by industry, with nearly 83% of education jobs and 63% of broad professional services jobs able to be done from home – while only 3.5% of jobs in the accommodation industry could be done from home.
Caveats: These calculations are indicative only, as it assumes a similar ICT take-up rate between NZ and USA. However, this framework provides an incentive for a deeper research that can be done using NZ administrative and Census data.