Working Papers

clay-banks-c2a0TydMlAs-unsplash (1)_edited.jpg

Constructing real time estimates of Australian consumer
spending using bank transactions

Matthew Elias

Bank transactions have been increasingly used by researchers to generate timely insights about consumer spending. The e61 Institute currently maintains the Spendtracker, a weekly tracker of consumer spending in Australia using bank transaction data supplied by
credit bureau illion. This technical document describes the methodological construction of the Spendtracker and validates its performance as an indicator of consumer spending against other publicly available datasets in Australia. After appropriate data cleaning and transformation we find that our estimates from Spendtracker co-move closely with other published data sources on consumer spending from official statistics in close to real-time.

Creative Work_edited.jpg

COVID and firm productivity

Dan Andrews, Andrew Charlton, Angus Moore

In typical recessions, less productive firms shrink by more than their more productive counterparts. This “cleansing” process of recessions reallocates jobs from less- to more-productive firms, in the process improving aggregate productivity. This project uses data on hundreds of thousands of small businesses in Australia, New Zealand, and the United Kingdom to look at whether this reallocation occurred during COVID. This is important, because COVID was not a typical recession. Firms were forced to close uniformly, regardless of their productivity.


Gig economy labour market

Jeff Borland, Andrew Charlton, Amit Singh, Oliver Alexander

We investigate the labour market for Uber drivers in Australia using administrative and survey data. Uber drivers’ total hours of work and driving schedules exhibit substantial heterogeneity and week-to-week variation; which appear to mainly reflect drivers’ preferences. We identify several pathways to driving with Uber, associated with different effects on income and job satisfaction.  Variability in earnings between drivers depends primarily on differences in the number of trips per hour – which in turn is related to job tenure, time and location of driving, and the proportion of offered trips accepted by drivers.


International comparison of wage subsidies

Steve Hamilton, Andrew Charlton, Angus Moore

JobKeeper was the largest government support package introduced by the Australian government in response to the COVID crisis. It was a wage subsidy, designed to help firms forced to close by COVID-19 retain their staff, and prevent the severing of these job matches for businesses. This project uses data on hundreds of thousands of Australian small businesses and their employees to examine what impact the policy had. We look at the types of firms that applied for it and their behaviour; the impact it had on employee retention and wages; and what other impacts it had on firm performance – such as retention of other non-eligible employees, sales growth, cashflow, etc.

Get in Touch
Seniors Laughing_edited.jpg

Pension withdrawal as stimulus

Steven Hamilton, Andrew Charlton, Geoffrey Liu

Pensions represent an alternative funding source for cash stimulus during recessions, enabling governments to overcome constraints on public debt. During the COVID- 19 crisis, the Australian Government provided cash stimulus of up to AUD$20,000 (US$15,000) per person via early pension withdrawals, with a quarter of the working-age population withdrawing 3% of GDP. Using detailed bank account transaction data, we find a marginal propensity to consume (MPC) out of these withdrawals of 0.6, with the effect gradually falling over a 6-week period. This suggests an immediate GDP impact of just under 2%. Among those who took up the policy, treatment effects increased substantially with low liquidity, income loss and welfare recipient status, suggesting even tighter targeting would have raised the MPC.

Get in Touch