Research Interests
Climate Risk, Housing and Real Estate Economics, Urban and Regional Economics, Credit Risk Modelling, Index Numbers and Economic Measurement, Causal Modelling Methods.
Current Work
Funded Research:
ARC Grant (2022-2025): Measuring the Commercial Real Estate Sector in Australia [with Alicia Rambaldi and Robert Hill]
The project will develop method for constructing price indexes for commercial property. A key output of the project will be commercial real price indexes for a range of segments and regions across Australia. We are also investigating the impact of NABERS energy efficiency ratings on the commercial real estate market.
Climate Risk:
Climate Change and its Impact on Home Insurance Uptake in Australia [with Trinh Le and Ummul Ruthbah] (Forthcoming, Ecological Economics)
Climate change is impacting the frequency and severity of physical risks such as wildfire, flood, cyclone and extreme precipitation. This is changing both the costs and benefits of home insurance purchase for households—leading to uncertainty about future insurance uptake rates under climate change. We use disaggregated data for Australia on the likelihood of climate hazards, and household panel data, to model the relationship between insurance uptake and exposure to climate physical risks. This model is then used to project insurance purchase out to 2100 under three climate scenarios. We find that exposure to climate risks does influence the likelihood of purchasing insurance. While for most hazards higher risk reduces the likelihood of purchase, there are exceptions, and the relationship is often complex. Projections from our model indicate that insurance coverage is likely to decline relatively modestly, by around one percentage point across Australia from 2000 to 2100 as a result of climate change. However, we find there is significant variation across regions.
Climate Physical Risks and Mortgage Default in the US [with Dieter Brand and Joan Tan]
As the climate changes, natural disasters—such as floods, hurricanes, and wildfires—are likely to increase in frequency and severity. These climate physical risks have broad implications including for borrowers and banks. This study explores how extreme climate events in the US have impacted the ability of borrowers to pay their mortgages. We use detailed data at the MSA level from 2000 to 2023 on mortgage performance, property damage caused by hazard events, and the local economy, to construct impulse response functions to climate events. Our findings show significant impacts on severe delinquency, default, and prepayment as well as the local labour and housing markets.
Farm Prices and Climate Change [with ABARES]
This paper uses a large dataset to explore how changes in the climate have changed Australian farmland prices over the past 40 years. Using this relationship, we project how farmland prices are likely to change as climate change evolves out to 2100.
Index Numbers and Economic Measurement:
The Welfare Effects of New and Disappearing Goods (Revise & Resubmit)
This study proposes a new approach to quantifying the benefits of product variety to consumers using the flexible quadratic mean-of-order-r utility function. This allows for more flexible substitution patterns between product varieties than existing approaches, such as CES. A novel regularization approach is used to estimate this functional form at large scale—with many thousands of product varieties. The method is applied to supermarket scanner data from New York for 10 product categories from 2006 to 2020. Compared with the widely used CES approach, we find smaller gains to consumers from changes in product variety over time.
Reconciling Price Indices of Different Frequencies: An Application to the Housing Market [with Michael Scholz, Alicia Rambaldi and Robert Hill] (Under Review)
Central banks and government agencies need reliable higher frequency economic statistics to make informed decisions on monetary policy, fiscal policy and regulatory interventions. In this paper, we develop a least-squares approach for reconciling price indices of different frequencies (e.g., monthly, quarterly and annual). This reconciliation approach reduces the inconsistencies across frequencies and hence improves reliability. We demonstrate this improvement using a novel criterion applied to two housing related empirical examples. The first application uses indices for Sydney, Australia that we construct ourselves from micro-level data. The second uses off-the-shelf US indices. In both cases, the gains are largest for higher frequency indices which leverage on the greater stability of their lower frequency counterparts.
Measuring Inflation at High Frequency With Consumer Stockpiling [with Alicia Rambaldi and Robert Hill]
Electronic point of sale data has obvious benefits for statistical agencies. It records quantities, as well as prices, of products sold and it provides a census of transactions, rather than a sample. Scanner data is now used in constructing the Consumer Price Index (CPI) in a number of countries including; Australia, Belgium, Denmark, Iceland, the Netherlands, New Zealand, Norway, Sweden and Switzerland (ABS, 2016). In this paper we propose an approach for constructing high frequency measures of consumer price changes. The main challenge is stockpiling behaviour by consumers. Stockpiling can cause drift in chained price indices even when superlative formulas are used. We estimate the gap required between weeks being compared to avoid the stockpiling problem for various products and then based on this information propose a weighted GEKS approach for estimating weekly price indices that are free from drift. The paper compiles weekly indices of price changes for various supermarket products for a sample of US cities over the period 2006-2020.