New Zealand decision makers now have access to more frequent unemployment statistics thanks to a new tool.
The Unemployment Rate – Nowcast model, called UR-NOW, predicts the unemployment rate each month, published here, to supplement the official quarterly figures.
AUT Professor Tim Maloney, also Chief Economist at the Ministry for Social Development, and Principal Advisor David Rea
at the Ministry of Social Development, made the tool and described its development in a working paper.
The official unemployment rate is based on Stats NZ’s Household Labour Force Survey over a three-month period, so there
is a natural delay of up to four months while that information is collected, analysed and published.
It is an important economic and social indicator with implications for decision-making in many areas of the economy.
Many other OECD countries, including the United Kingdom, the United States of America, and Australia, run labour-force
surveys that provide monthly estimates of unemployment.
“The speed and potential scale of the economic downturn in 2020 highlighted the importance of timely information about
the labour market,” Professor Maloney says.
“Most of the time the slight lag in the official unemployment rate is not a big problem, because economic conditions are
slow to evolve.
“But as the current recession shows, when the environment is changing rapidly and in ways that are difficult to
forecast, being able to monitor real-time labour-market conditions is immensely beneficial.”
The UR-NOW tool uses constantly updated data about the receipt of benefits and the number of filled jobs to deliver a
predicted monthly unemployment rate.
Those predictions are intended to supplement rather than replace the existing survey-based figures.
The regular survey results allow regular re-estimation and validation of the tool, and also provide an independent check
on unusual HLFS results.
Professor Maloney says the researchers have been encouraged by initial interest from university academics and various
government ministries who could use the tool in their monitoring, policy development and advisor capacity.
“There is plenty of scope for future improvements to UR-NOW, including more frequent predictions, incorporating new
administrative data from other sources, predictions of additional labour market indicators, and predictions of
unemployment rates for specific population sub-groups and geographical areas.”
UR-NOW predictions will appear on this page of the Centre for Social Data Analytics at AUT, usually at the end of the week after the month finishes.