Unveiling the Australian Housing Sector Crisis: An Empirical Investigation
Aneesh Agarwal
Research Questions & Data
Research Questions
What is the nature of the relationship between housing credit and the housing price index in Australia? How do fluctuations in housing credit impact changes in house prices over time?
How do trends in housing debt, disposable income, and assets compare over the examined period? What insights do these comparisons offer into individuals’ financial health and the stability of the housing market?
How effective is the use of the inter-bank cash rate in combating rising price indexes in Australia? What impact do changes in the inter-bank cash rate have on housing affordability and market stability?
Data Details
The analysis utilizes an observational data set spanning from the first quarter of 2010 to the first quarter of 2024, compiled from data provided by both the Reserve Bank of Australia and the Australian Bureau of Statisitcs. The data has been prepared by combining multiple data sets containing information on:
Residential Property Price Indexes: Eight Capital Cities (RBA)
Monthly Consumer Price Index indicator (RBA)
Growth in Selected Financial Aggregates (ABS)
Household Finances – Selected Ratios (ABS)
Interest Rates and Yields – Money Market (ABS)
Consumer Price Inflation – Expenditure Groups (ABS)
The data used in the analysis contains the following variables:
data_dictionary <- read_excel("Analysis Data/data-dictionary.xlsx")
data_dictionary |> select(NAME, DESCRIPTION) |>
kable(format = "pipe")| NAME | DESCRIPTION |
|---|---|
| Quarter | Year & Quarter |
| DGFACHM | Housing Credit; Quaterly growth |
| DGFACIHM | Investor housing Credit; Monthly growth |
| BHFHDHA | Ratio of housing debt to housing assets |
| BHFDDIH | Ratio of housing debt to annualised household disposable income |
| GCPIHOYP | Housing price index (% change) |
| FIRMMCRIHM | Highest Interbank Overnight Cash Rate; Quaterly average |
| A83728455L | Weighted Average of Residential Property Price Index of eight capital cities ; |
| A83728458V | Weighted Average of Residential Property Price of eight capital cities (% change) |
| A128485865L | Index of Housing |
| A128473449C | Index of Rent |
| A128475405L | Index of New dwelling purchase by owner-occupiers |
Additional Information on the variables can be obtained from data-dictionary.xlsx file.
The data obtained and combined from ABS and RBA is provided under Creative Commons Attribution 4.0 International License (CC BY 4.0 Licence).
Remember
- data-dictionary.csv
Blog post
Introduction
The housing discourse in Australia has shifted from casual discussions to a crisis narrative, highlighting the pressing issue of unaffordability for both buyers and renters. Formerly seen as a problem affecting the disadvantaged, the housing crisis now encompasses a broader spectrum of society, including essential workers and homeowners concerned about future generations’ access to housing (Parsell, Kuskoff, and Reddel 2023). Factors such as decreased investment in social housing, policy decisions favoring speculation, and a significant gap between supply and demand have intensified the crisis (Karp 2024). Additionally, the return of international students and migrants post-COVID has exacerbated housing shortages (The Salvation Army Australia n.d.).
p1 <- master |>
ggplot(aes(x = Quarter, y = GCPIHOYP)) +
geom_line(color = "grey") +
geom_point(color = "white", size = 2) +
labs(title = "Movement in Housing Price Index since Q1 2016" ,y = "Housing Price Index", x = "Time Period") +
theme_modern_rc()+
theme(plot.title = element_text(size = 10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8))
p2 <- master |> drop_na() |>
ggplot(aes(x = Quarter)) +
geom_line(aes(y = A128485865L, color = "Housing Index")) +
geom_point(aes(y = A128485865L)) +
geom_line(aes(y = A128473449C, color = "Rent Index")) +
geom_point(aes(y = A128473449C)) +
geom_line(aes(y = A128475405L, color = "New Dwellings Index")) +
geom_point(aes(y = A128475405L)) +
labs(title = "Index of Rent, Housing & New Dwellings", y = "Index", x = "Time Period", color = "Legend") +
theme_modern_rc()+
theme(plot.title = element_text(size = 10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
p3 <- master |>
select(Quarter, A83728455L) |>
drop_na() |>
ggplot(aes(x = Quarter)) +
geom_line(aes(y = A83728455L)) +
geom_point(aes(y = A83728455L)) +
labs(title = "Weighted Residential Property Price Index of Eight Capital Cities", y = "Index", x = "Time Period") +
theme_modern_rc()+
theme(plot.title = element_text(size = 8),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
p4 <- master |>
select(Quarter, A83728458V) |>
drop_na() |>
ggplot(aes(x = Quarter)) +
geom_line(aes(y = A83728458V)) +
geom_point(aes(y = A83728458V)) +
labs(title = "Change in Weighted Residential Property Price Index of Eight Capital Cities", y = "Percentage Change", x = "Time Period") +
theme_modern_rc()+
theme(plot.title = element_text(size = 8),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
(p1+p2)/(p3+p4)
Figure 1: Affordability Trends in Housing Market of Australia
This study aims to empirically examine Australia’s housing crisis, analyzing affordability, supply, and demand trends (see Figure 1). Understanding these dynamics is crucial for policymakers and stakeholders to develop effective solutions, ensuring equitable access to secure and affordable housing for all Australians.
Motivation
As a student of economics and an international migrant, my personal experiences and academic background fuel a deep-seated motivation to explore and address Australia’s housing crisis. Having witnessed firsthand the challenges faced by migrants in securing affordable housing, I am acutely aware of the profound impact housing policies can have on individuals and communities. Moreover, my studies in economics have equipped me with analytical tools to understand the intricate interplay between market dynamics, policy decisions, and societal outcomes in the realm of housing. By delving into empirical research on housing affordability, supply, and demand, I aim to contribute to a deeper understanding of the systemic issues at play and identify potential avenues for policy intervention.
Data
The analysis utilizes a comprehensive data set spanning from the first quarter of 2010 to the first quarter of 2024, compiled from data provided by both the Reserve Bank of Australia and the Australian Bureau of Statisitcs. This combined data set encompasses key metrics including Housing Credit, Housing Debt, Housing Price Index, Rent Price Index, Residential Price Index as well as Inter-Bank Cash Rate. By leveraging this rich and diverse data set, the analysis aims to provide a nuanced understanding of the dynamics shaping Australia’s housing market over the examined period.
Analysis
Investigating the Relationship Between Housing Credit & Price Index
In Australia, it’s no secret that the housing market is a dynamic beast, where the availability of housing credit and the price of homes are deeply intertwined (Fazekas 2023). When there’s a surge in housing credit, you can usually expect house prices to follow suit, and vice versa. But here’s the twist: these changes aren’t always long-lasting. Especially when it comes to investors, shifts in housing credit can cause quick spikes or drops in house prices, but they tend to be temporary (Saunders and Tulip 2019).
ardl_plot_1 <- master |>
select(Quarter, GCPIHOYP, DGFACIHM, DGFACHM) |>
drop_na() |>
ggplot(aes(x = Quarter)) +
geom_line(aes(y = GCPIHOYP)) +
labs(title = "Housing Price Index", x = "Time Period") +
theme_modern_rc() +
theme(plot.title = element_text(size = 10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
ardl_plot_2 <- master |>
select(Quarter, GCPIHOYP, DGFACIHM, DGFACHM) |> ggplot(aes(x = Quarter)) +
geom_line(aes(y = DGFACHM)) +
labs(title = "Housing Credit", x = "Time Period") +
theme_modern_rc() +
theme(plot.title = element_text(size = 10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
ardl_plot_3 <- master |>
select(Quarter, GCPIHOYP, DGFACIHM, DGFACHM) |> ggplot(aes(x = Quarter)) +
geom_line(aes(y = DGFACIHM)) +
labs(title = "Investor Housing Credit", x = "Time Period") +
theme_modern_rc() +
theme(plot.title = element_text(size = 10),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
ardl_plot_1 / (ardl_plot_2+ardl_plot_3)
Figure 2: Time Series Plot of Housing Price Index, Housing and Investor Credit
Figure 2 illustrates the trajectory of the Housing Price Index, Housing Credit, and Investor Housing Credit spanning from the first quarter of 2010 to the fourth quarter of 2023. Notably, a discernible downward trend is evident in the data leading up to the first quarter of 2020, suggesting a period characterized by subdued credit and pricing dynamics. However, the advent of the COVID-19 pandemic ushered in a notable shift, precipitating a sharp upturn in all variables, marking a pronounced departure from the established trend.
Furthermore, it’s worth noting that the ascent in the price index is preceded by corresponding increases in both housing credit and investor housing credit values. This observation suggests a certain degree of interdependence among these variables, where subsequent values are influenced by their preceding counterparts.
In our analysis, we’re diving deep into the housing market dynamics using a sophisticated tool called the Auto-Regressive Distributed Lag (ARDL) Regression Model. This model helps us understand how today’s housing credit, investor credit, and housing prices are not just influenced by their current values, but also by their past trends.
Imagine it like this: the ARDL model lets us peek into the rear-view mirror of the housing market. We’re not just looking at what’s happening right now; we’re also considering how past movements in housing credit and investor credit have shaped today’s housing prices, and vice-versa.
By harnessing the power of this model, we’re able to uncover the subtle connections and inter-dependencies between these key variables. This deeper insight helps us make more informed predictions and decisions in the ever-evolving landscape of the housing market.
model1 <- auto_ardl(data = master, formula = GCPIHOYP ~ DGFACHM + DGFACIHM, max_order = 5)
summary_model1 <- tidy(model1$best_model)
summary_model1[,1:2] |> kable(format = "pipe", align = "lc", caption = "Table 1: Coefficients of Distributed Lag Model")| term | estimate |
|---|---|
| (Intercept) | -0.4638028 |
| L(GCPIHOYP, 1) | 1.2318026 |
| L(GCPIHOYP, 2) | 0.0565251 |
| L(GCPIHOYP, 3) | -0.3932318 |
| DGFACHM | 1.9628005 |
| DGFACIHM | -0.3535238 |
| L(DGFACIHM, 1) | 2.2217303 |
| L(DGFACIHM, 2) | -2.0181007 |
Table 1 shows the coefficients of the ARDL model run on our data. Let’s break down the interpretation of each coefficient in simpler terms:
Intercept: This is our starting point. It’s like the base level of housing prices when nothing else is changing.
L(GCPIHOYP, 1): If housing prices went up by $1 in the last quarter, we’d expect them to go up by roughly $1.23 in the current quarter.
L(GCPIHOYP, 2): If housing prices increased by $1 two quarters ago, we’d expect them to rise by about $0.06 in the current quarter.
L(GCPIHOYP, 3): If housing prices went up by $1 three quarters ago, we’d anticipate them to decrease by around $0.39 in the current quarter.
DGFACHM: For every extra dollar of housing credit taken out this quarter, we’d anticipate housing prices to go up by roughly $1.96.
DGFACIHM: With every additional dollar of investor housing credit this quarter, we’d expect housing prices to drop by approximately $0.35.
L(DGFACIHM, 1): If investor housing credit increased by $1 last quarter, we’d predict housing prices to increase by about $2.22 this quarter.
L(DGFACIHM, 2): Similarly, if investor housing credit went up by $1 two quarters ago, we’d foresee housing prices to decrease by roughly $2.02 this quarter.
These interpretations provide a clear picture of how changes in housing credit, investor credit, and past trends in housing prices influence the current state of the housing market. The ARDL model findings tell us a few important things about the housing market. When we look at past housing prices, there’s a bit of a trend: if prices went up last quarter, they’re likely to keep going up this quarter too, but if they shot up two or three quarters ago, they might drop a bit now. When more people borrow money for houses, prices tend to go up, which makes sense. But if investors borrow more money, prices could actually drop, which is a bit surprising. It seems that when investors borrow more last quarter, prices go up this quarter, but if they borrowed a lot two quarters ago, prices might go down now. Overall, it’s a bit of a mix between how much people are borrowing and what happened to prices in the past that determines what’s happening with house prices now.
Comparing Housing Debt, Disposable Income and Assets
Housing Debt to Income measures how much of a person’s income goes towards paying housing-related debts. A higher value indicates a larger portion of income being used to service housing debts, which could indicate financial strain for individuals or potentially higher levels of debt in the housing market overall. Housing Debt to Assets gauges the proportion of a person’s assets that are tied up in housing debt. A higher value suggests that a larger portion of assets is leveraged to finance housing-related expenses, which could indicate a riskier financial position for individuals or potentially higher levels of systemic risk in the housing market.
dia_data <- master |> select(Quarter, BHFHDHA, BHFDDIH) |> drop_na()
p5 <- dia_data |> ggplot(aes(x = Quarter)) +
geom_line(aes(y = BHFDDIH)) +
labs(title = "Housing Debt to Income", x = "Time Period", color = "Legend") +
theme_modern_rc() +
theme(plot.title = element_text(size = 12),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
p6 <- dia_data |> ggplot(aes(x = Quarter)) +
geom_line(aes(y = BHFHDHA)) +
labs(title = "Housing Debt to Housing Assets", x = "Time Period", color = "Legend") +
theme_modern_rc() +
theme(plot.title = element_text(size = 12),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
p5+p6
Figure 3: Housing Debt, Disposible Income and Assets
Figure 3 provides data for two key metrics: Housing Debt to Income and Housing Debt to Assets over several quarters from 2010 to 2024. These metrics are vital for understanding individuals’ financial health and the housing market’s stability.
From 2010 to 2023, there is a general increasing trend in Housing Debt to Income, indicating that housing debt relative to income has been on the rise over this period. While there are some fluctuations quarter-to-quarter, the overall trend suggests a gradual increase in the burden of housing debt on individuals’ incomes. This trend could signify potential challenges in housing affordability or an increase in debt-financed housing purchases over time.
Housing Debt to Housing Assets shows a more varied trend compared to BHFDDIH, with fluctuations in values across different quarters. There are periods of increase and decrease in Housing Debt to Housing Assets, indicating shifts in the proportion of assets allocated to housing debt over time. The overall trend appears less consistent than that of Housing Debt to Income, suggesting fluctuations in the proportion of assets tied up in housing debt relative to total assets.
Combating Rising Price Indexes with Inter Bank Cash Rate
Central banks often use interest rate adjustments as a tool to achieve price stability and other macroeconomic objectives (Kearns 2022). The interbank cash rate holds significance beyond financial circles; it can indirectly impact the prices we encounter daily. Here’s how:
Influence on Borrowing and Spending: Central banks often adjust the interbank cash rate to regulate borrowing costs for banks. This, in turn, affects interest rates across the economy. Lower rates make borrowing cheaper, prompting increased spending by businesses and consumers. This heightened demand can push prices upward, contributing to inflation.
Impact on Investment and Consumer Behavior: Changes in interest rates influence investment decisions and consumer behavior. Lower rates encourage businesses to invest in growth, while consumers may be more inclined to make purchases using credit. This surge in demand can fuel price increases, reflecting in various price indices.
Exchange Rate Effects: Alterations in interest rates can also influence exchange rates. A decrease in the interbank cash rate may lead to a depreciation of the domestic currency, making imports more expensive. This, in turn, can contribute to inflationary pressures by raising the prices of imported goods and services.
Market Expectations and Sentiment: Market participants closely monitor central bank decisions regarding interest rates for insights into future economic conditions. Expectations about interest rate movements can influence investor behavior and asset prices, indirectly affecting overall price levels.
While adjusting interbank rates may not directly remedy price indices, they play a crucial role in shaping economic conditions that impact inflation and price stability. Understanding these connections helps us grasp the broader implications of monetary policy decisions on our everyday lives.
master |> filter(Quarter >= "2016-1") |>
select(Quarter, GCPIHOYP, FIRMMCRIHM) |>
drop_na() |>
ggplot(aes(x = Quarter)) +
geom_line(aes(y = GCPIHOYP, color = "Housing Price Index")) +
geom_line(aes(y = FIRMMCRIHM, color = "Highest Interbank Cash Rate")) +
labs(y= "" ,x = "Time Period", color = "Legend") +
theme_modern_rc() +
theme(plot.title = element_text(size = 12),
axis.text.x = element_text(angle = 90, vjust = 0.5, hjust=1, size = 8),
axis.text.y = element_text(size = 8),
axis.title.x = element_text(vjust = -3.5))
Figure 4: Examining the Trends in Cash Rate and Price Index
After examining Figure 4, it becomes evident that a substantial shift occurred following the first quarter of 2022. Specifically, the Reserve Bank of Australia (RBA) initiated a series of significant increases in the Inter-bank cash rate. This monetary policy maneuver by the RBA, aimed at managing inflation and economic stability, had a pronounced impact on the housing market.
As a direct consequence of the rise in the Inter-bank cash rate, there was a discernible downward trajectory observed in the housing price index. This decline reflects the tightening of borrowing conditions and the increased cost of mortgage financing. Homebuyers and property investors, facing higher borrowing costs, likely became more cautious, leading to reduced demand in the housing market.
Conclusion
Australia’s housing crisis has become a pressing issue affecting a wide range of individuals, from essential workers to homeowners concerned about future generations’ access to housing. Factors such as decreased investment in social housing, speculative policy decisions, and a significant supply-demand gap have intensified this crisis, exacerbated further by the return of international students and migrants post-COVID. Our analysis, covering data from 2010 to 2024, reveals intricate dynamics within the housing market. We’ve uncovered the close relationship between housing credit availability, investor behavior, and housing prices. Shifts in housing credit, particularly driven by investors, can lead to rapid price fluctuations, showcasing the market’s sensitivity to financial factors. Moreover, the impact of monetary policy, notably changes in the Inter-bank cash rate, is significant. The Reserve Bank of Australia’s decision to increase interest rates post-2022 Quarter 1 resulted in a notable downturn in the housing price index, highlighting the effectiveness of monetary policy in influencing market conditions.
In addressing the housing crisis, a multifaceted approach is essential. Policy interventions aimed at increasing housing supply, curbing speculative investment, and ensuring affordability are crucial. Additionally, careful consideration of monetary policy decisions’ implications for the housing market is vital for achieving sustainable and inclusive housing outcomes. As we navigate these challenges, leveraging insights from our analysis to inform evidence-based policy solutions and collaborative efforts is crucial for fostering a more equitable and resilient housing system for all Australians.
References
Behind the Scenes
Challenges Faced During Data Collection
Data Extraction
The data within the raw data sets was found to be inadequately structured, as metadata was commingled with the substantive data. As a result, it became imperative to disentangle these components and subsequently arrange them in a manner conducive to rigorous analysis, as depicted in the figure below.
Furthermore, the study necessitated the amalgamation of data sourced from six distinct datasets, each requiring individual formatting. Additionally, the process entailed the selective extraction of a subset of variables from each data set, with the exclusion of superfluous variables proving to be a repetitive and cumbersome endeavor.
Data Interpolation
A considerable portion of the variables within the data set exhibited a monthly frequency, whereas others were recorded quarterly. To ensure uniformity in both analysis and structure, it became necessary to interpolate the data to a quarterly frequency. This process facilitated a harmonized temporal framework for comprehensive analysis and interpretation.
Challenges Faced During Data Analysis
Missing Observations
A notable limitation of the study stems from the incomplete updating of information within the datasets by the respective authorities. Consequently, the analysis encountered constraints, as not all variables could be comprehensively assessed. Instead, the examination proceeded incrementally, addressing the available information in fragments rather than as a cohesive whole.
Exclusion of Variables
In light of the restricted scope and available resources, certain variables initially incorporated into the amalgamated data set had to be subsequently excluded. This measure was undertaken to streamline the analysis process and optimize resource allocation, thereby ensuring a more focused examination of the pertinent variables within the data set.
Scope for Future Research
The study, constrained by limited familiarity with R libraries tailored for econometric analysis, as well as time and resource constraints, predominantly relies on graphical analysis. Furthermore, it emphasizes a singular remedy for addressing housing prices amidst a plethora of alternative solutions. It is worth noting that future studies, afforded the opportunity, can adopt a more exhaustive and detailed methodology. Such endeavors could leverage a broader array of sophisticated packages and encompass a more extensive range of empirical solutions, thereby enhancing the depth and breadth of analysis within this domain.