Property Infrastructure: Why Big Projects Fail Investors - Expert Busting Series

Investors spend a lot of time looking into rail lines, hospital builds, and highway upgrades. The belief is that big projects drive big growth.

Infrastructure amenities attract buyers. Construction brings jobs. Jobs bring people. People drive demand.

When you test this against historical data, the relationship either disappears or runs in the wrong direction.

Infrastructure research is one of the biggest time sinks in property investing. And it is largely unnecessary.

Testing It With Real Data

The ABS Engineering Construction Activity dataset tracks public spending on:

Roads
Bridges
Railways
Pipelines
Harbours
Water storage
Electricity
And more civil infrastructure projects

It only captures data at the state level. But it answers the most fundamental question first: does the relationship between infrastructure spending and capital growth exist at all?

Every state and territory was measured each quarter over 25 years. Engineering construction activity per capita, adjusted for inflation, was plotted against the capital growth that followed over the next 3 years.

If infrastructure drives growth, high-spending states should show high following growth. The trend line should slope upward.

The scatter plot below shows every observation. Look at the trend line before reading further.

Scatter plot titled Engineering Construction Activity Per Capita Inflation-Adjusted versus 3-Year House Price Growth. Each white circle represents a state or territory's engineering construction activity for a quarter over 25 years. The horizontal X axis shows engineering construction activity per capita in dollars from $0 to $5,000. The vertical Y axis shows 3-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line runs almost perfectly horizontal across the chart, indicating no meaningful relationship between infrastructure spending and following capital growth.

The trend line was almost dead flat. No relationship between engineering construction activity and 3-year capital growth.

Changing the Measure. Same Result.

The next test measured the change in engineering construction activity over 3 years. Does a surge in spending predict a surge in growth?

No meaningful relationship in the expected direction.

The chart below shows the same test using change in spending instead of raw dollars. Different measure. Same result. The trend line is flat.

Scatter plot titled 3-Year Engineering Construction Activity Per Capita Inflation-Adjusted Growth versus 3-Year House Price Growth. The horizontal X axis shows 3-year change in engineering construction activity per capita from minus 50 to 90 percent per annum. The vertical Y axis shows 3-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line runs almost perfectly horizontal across the chart indicating no meaningful relationship between the change in infrastructure spending and following capital growth.

The timeframe was extended to 6 years. A correlation appeared. But it sloped the wrong way.

The chart below shows both axes measuring 6-year change. Before reading the numbers, look at the direction of the trend line.

Scatter plot titled 6-Year Engineering Construction Activity Per Capita Inflation-Adjusted Growth versus 6-Year House Price Growth. The horizontal X axis shows 6-year change in engineering construction activity per capita from minus 50 to 90 percent per annum. The vertical Y axis shows 6-year house price growth from minus 10 to 30 percent per annum. The dotted white trend line slopes downward from left to right, indicating that higher engineering construction activity growth was followed by lower capital growth over the same 6-year period.

Higher engineering construction activity was followed by lower capital growth. Not higher.

The Relationship Runs Backwards

When the sequence was reversed, measuring house price growth first and engineering construction activity after, a positive correlation appeared.

Capital growth leads to infrastructure spending. Not the other way around.

State governments collect stamp duty based on property values. Higher prices fund new projects.
Local councils collect rates based on land values. Higher land values fund local infrastructure.
Federal contributions are tied to state co-funding. State revenue ties back to land values.

When federal and state contributions were removed and only local council spending was analysed, the correlation strengthened. Council budgets are almost entirely dependent on rates. Rates are almost entirely dependent on land values.

Infrastructure does not push up prices. Rising prices fund infrastructure.

The chart below shows what happens when the sequence is flipped. House price growth comes first. Engineering construction activity follows. The trend line now slopes upward.

Scatter plot titled 6-Year House Price Growth versus 6-Year Engineering Construction Activity Per Capita Inflation-Adjusted Growth. The horizontal X axis shows 6-year engineering construction activity growth percentage per annum from minus 50 to 90. The vertical Y axis shows 6-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line slopes upward from left to right, showing that higher house price growth was followed by higher engineering construction activity. The correlation is positive and runs in the opposite direction to what infrastructure-led growth theory predicts.

A Second Dataset. The Same Finding.

The ABS Building Activity Non-Residential dataset covers:

Hospitals
Universities
Offices
Transport buildings
Education buildings
And more

These are different types of projects to Engineering Construction Activity, but still of keen interest to investors.

The chart below shows non-residential building activity per capita plotted against 3-year capital growth. Look at the trend line.

Scatter plot titled Building Activity Non-Residential Per Capita Inflation-Adjusted Dollar versus 3-Year House Price Growth. The horizontal X axis shows non-residential building activity per capita in dollars from 0 to 500. The vertical Y axis shows 3-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line runs almost perfectly horizontal across the chart indicating no meaningful relationship between non-residential building activity and following capital growth.

The trend line against 3-year capital growth was flat.

Change in non-residential building activity over 6 years showed the same inverse pattern. Higher building activity. Lower following growth.

The chart below shows that inverse pattern. Both axes measure 6-year change. The trend line falls to the right.

Scatter plot titled 6-Year Building Activity Non-Residential Per Capita Inflation-Adjusted Growth versus 6-Year House Price Growth. The horizontal X axis shows 6-year change in non-residential building activity percentage per annum from minus 50 to 90. The vertical Y axis shows 6-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line slopes downward from left to right indicating that higher non-residential building activity growth was followed by lower capital growth over the same 6-year period.

When the sequence was reversed, house price growth again preceded building activity.

The chart below shows the reversed sequence for the non-residential building dataset. House price growth leads. Building activity follows. The same pattern holds across both datasets.

Scatter plot titled 2-Year House Price Growth versus 2-Year Building Activity Non-Residential Per Capita Inflation-Adjusted Growth. The horizontal X axis shows 2-year change in non-residential building activity percentage per annum from minus 50 to 90. The vertical Y axis shows 2-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line slopes upward from left to right, confirming that house price growth preceded non-residential building activity across the dataset.

Two separate datasets. The same result both times.

One Weak Signal

The ABS Private Sector Capital Expenditure dataset covers mining, manufacturing, construction, retail, finance, and health. This spending is not dependent on rising land values.

A 3-year change in private capital expenditure showed a weak positive correlation with 2-year capital growth that followed. The trend line climbed slightly. The direction was correct.

The chart below shows that weak signal. The trend line points in the right direction. But look at how scattered the dots are. The relationship exists. But it is not reliable enough to act on.

Scatter plot titled 3-Year Private Capital Expenditure Per Capita Inflation-Adjusted Growth versus 2-Year House Price Growth. The horizontal X axis shows 3-year change in private capital expenditure percentage per annum from minus 50 to 70. The vertical Y axis shows 2-year house price growth percentage per annum from minus 10 to 30. The dotted white trend line slopes slightly upward from left to right indicating a weak positive correlation. The dots are widely scattered with no obvious clustering pattern confirming the correlation is weak.

This was the strongest signal found across all three datasets. It was not strong enough to build a strategy on.

120 Projects. One in Ten Made a Difference.

Individual projects were examined next. Each was assessed by comparing impacted markets against control markets before and after opening.

120 projects were analysed. Some results from individual projects follow.

Brisbane Airport New Runway
2012-2020
$1.1b
Pre-opening impacted 14.3% versus control 13.8%
Post-opening impacted 54.3% versus control 55.0%
No clear uplift.

The table below shows the full suburb-level breakdown for Brisbane Airport. Impacted markets on the left. Control markets on the right. Pre-opening figures top. Post-opening figures bottom. The numbers tell the same story across every suburb.

Data table titled Brisbane Airport New Runway 2012 to 2020 showing impacted and control markets across two time periods. The top left panel shows impacted markets including Hendra, Nundah, Banyo, Boondall, and Northgate with median house prices and percentage changes for the 4-year period to July 2020. Average change for impacted markets was 14.3% or 3.6% per annum. The top right panel shows control markets including Stafford, Chermside West, Mount Gravatt, Carina, Moorooka, and Brisbane LGA with an average change of 13.8% or 3.5% per annum. The bottom left panel shows impacted markets for the 4-year period to July 2024 with an average change of 54.3% or 13.6% per annum. The bottom right panel shows control markets for the same period with an average change of 55.0% or 13.8% per annum. Impacted markets did not outperform control markets in either period.

Albion Park Rail Bypass
2019-2021
$630m
Impacted markets weaker than controls before and after opening
No clear uplift.

The table below shows the suburb-level breakdown for the Albion Park Rail Bypass. Impacted markets on the left. Control markets on the right. The control markets outperformed in both periods.

Data table titled Albion Park Rail Bypass 2019 to 2021 showing impacted and control markets across two time periods. The top left panel shows impacted markets including Albion Park Rail, Albion Park, Oak Flats, Yallah, and Dapto with median house prices and percentage changes for the 4-year period to October 2021. Average change for impacted markets was 21.3% or 5.3% per annum. The top right panel shows control markets including Woonona, Warrawong, Warilla, Mount Warrigal, and Barrack Heights with an average change of 25.0% or 6.2% per annum. The bottom left panel shows impacted markets for the 4-year period to October 2025 with an average change of 20.6% or 5.1% per annum. The bottom right panel shows control markets for the same period with an average change of 24.7% or 6.2% per annum. Control markets outperformed impacted markets in both periods indicating no clear bypass uplift.

Bruce Highway Upgrade
2013-2020
$9.57b
Impacted markets averaged 5.6% pre-completion versus control 7.4%
Markets recovered post-completion
The upgrade cannot be credited.

The table below shows the LGA-level breakdown for the Bruce Highway Upgrade. Five impacted LGAs on the left. Five control LGAs on the right. Pre-completion figures top. Post-completion bottom. The impacted markets underperformed before completion and recovered after. The highway did not drive that recovery.

Data table titled Bruce Highway 2013 to 2020 showing impacted and control markets across two time periods. The top left panel shows impacted LGAs including Gympie, Mackay, Townsville, Rockhampton, and Bundaberg with median house prices and percentage changes for the 5-year period to December 2020. Average change for impacted markets was 5.6% or 1.1% per annum. The top right panel shows control LGAs including Toowoomba, Western Downs, Southern Downs, Lockyer Valley, and South Burnett with an average change of 7.4% or 1.5% per annum. The bottom left panel shows impacted LGAs for the 4-year period to December 2024 with an average change of 72.4% or 18.1% per annum. The bottom right panel shows control LGAs for the same period with an average change of 69.5% or 17.4% per annum. Impacted markets underperformed controls pre-completion and recovered post-completion but the recovery aligns with broader Queensland market conditions not the highway upgrade.

New Bendigo Hospital
2016 and 2018
$630m
No growth at Stage 1 completion
Values rose from 2020
That coincided with COVID, not the hospital.

The chart below shows Bendigo property values against Melbourne from 2016 to 2023. Stage 1 completed in 2016. Stage 2 in 2018. Look at where the Bendigo line starts climbing.

Line chart titled New Bendigo Hospital 2016 and 2018 showing two lines from January 2016 to May 2023. The dark teal line tracks typical house values in Melbourne SUA. The green line tracks typical house values in Bendigo SUA. Both lines remain relatively flat from 2016 to early 2020. The Bendigo line begins rising sharply from mid-2020 coinciding with the COVID period not the hospital completion dates of 2016 and 2018. Melbourne follows a similar upward trajectory from the same period. No growth separation is visible around either hospital completion date.

Canberra New International Airport Terminal
2016
$2b
No discernible benefit compared to comparable cities.

The chart below shows Canberra property values alongside Sydney, Wollongong, Nowra, Goulburn, and other nearby cities from 2016 to 2021. The terminal opened in 2016. Look for any separation in the Canberra line after that date.

Line chart titled Canberra New International Airport Terminal 2016-09 showing multiple lines from September 2016 to September 2021. Lines track typical house values for Canberra-Queanbeyan, Sydney, Wollongong, Nowra-Bomaderry, Ulladulla, Goulburn, and Batemans Bay SUAs. The Canberra-Queanbeyan line shown in teal moves in line with comparable regional cities throughout the period. No visible separation or acceleration in Canberra values appears after the terminal opening in September 2016. All lines trend upward gradually and converge toward the end of the period.

WestConnex
2017
$1.6b
Sydney and western SA4 markets failed to reach 5% per annum in the years that followed.

The chart below shows Sydney SA4 property values from January 2016 to June 2020. WestConnex construction was underway from mid-2017. Look at how the lines move across all SA4 regions during and after construction.

Line chart titled West Connex Boosts Jobs 2017-06 showing five lines from January 2016 to June 2020. Lines track typical house values for Sydney SUA, Sydney Inner West SA4, Sydney Inner South West SA4, Sydney Baulkham Hills and Hawkesbury SA4, and Sydney Parramatta SA4. The Sydney SUA line sits highest throughout. All other lines cluster between approximately $736,000 and $1,200,000. No lines show meaningful acceleration during or after WestConnex construction from mid-2017. All lines remain relatively flat or decline slightly through 2018 and 2019 before stabilising in 2020.

After analysing 120 projects, only 1 in 10 showed any measurable impact on surrounding property values.

The exceptions shared one condition. Remote location. Large number of construction workers relative to local population. Project duration long enough to sustain rental demand. Once construction ended, demand subdued.

An Infrastructure Expert With 500 Reports

One revered expert has published infrastructure hotspot reports since 2007. Over 500 recommendations made across 13 years. They considered infrastructure to be the ultimate driver of residential property values.

Those 500 recommendations were tested against the benchmark. The benchmark was median house price growth across all Australian Significant Urban Areas.

Alpha measures how far above or below the benchmark each recommendation performed. Zero alpha means the same as random suburb selection. Negative alpha means underperformance.

The chart below shows the first recommendation from their 2007 report. Adelaide. The turquoise line is Adelaide growth per annum. The purple line is the national rate. Look at where Adelaide sits relative to the national line across the full period.

Line chart titled Capital Growth showing Adelaide versus Australia from October 2007 to October 2025. The turquoise line tracks Adelaide per annum growth rate starting from October 2007 when the report was published. The purple line tracks the national growth rate across the same period. Adelaide starts well above the national rate in 2007 and 2008 with growth approaching 20% per annum. Both lines converge and decline through 2009. Adelaide remains above the national rate through most of the period from 2010 to 2015. The lines sit close together from 2016 to 2022. Adelaide rises above the national rate again from 2022 onward finishing above the national line in 2025. The turquoise line is mostly above the purple line across the full period indicating Adelaide outperformed the national benchmark for the majority of the time since the recommendation.

The chart below shows the same data with the alpha shaded. The teal area shows when Adelaide outperformed the benchmark. The purple area shows when it fell behind. Adelaide was a genuine win. But one win does not prove anything.

Shaded area chart titled Capital Growth showing Adelaide versus Australia from November 2007 to October 2025. The teal shaded area represents periods when Adelaide growth exceeded the national benchmark. The purple shaded area represents periods when Adelaide growth fell below the national benchmark. A large teal area appears from 2007 to approximately 2015 showing sustained outperformance in the early years following the recommendation. The teal area narrows and disappears behind the purple shade from 2016 to 2022 indicating a period of underperformance. The teal area reappears from 2022 onward as Adelaide growth rises above the national rate again. The chart shows Adelaide outperformed for much of the period but with a significant gap of underperformance from 2016 to 2022.

The chart below shows the second recommendation from the same report. Cloncurry. Picked on the basis of five major mines and a $500 million copper project. The teal shaded area shows when Cloncurry outperformed the benchmark. The purple area shows when it fell behind. Look at how much of the chart is purple.

Shaded area chart titled Capital Growth showing Cloncurry versus Australia from October 2007 to October 2025. The teal shaded area represents periods when Cloncurry growth exceeded the national benchmark. The purple shaded area represents periods when Cloncurry growth fell below the national benchmark. A brief teal area appears in 2007 to 2008. The teal line then drops sharply into negative growth territory from 2009 to approximately 2016 with the shaded area turning fully purple. The Cloncurry line recovers slightly from 2016 but remains mostly below or near the benchmark until 2025. The chart shows Cloncurry underperformed the national growth rate for the majority of the period following the recommendation.

The chart below shows the alpha (performance above the nation) for every growth period from 1 to 18 years. A bar to the right means above benchmark (outperformance). A bar to the left means below (underperformance).

Horizontal bar chart titled Expert's Alpha 2007 to 2020. The vertical Y axis shows growth periods from 1 year at the bottom to 18 years at the top. The horizontal X axis shows alpha as percentage above the broader market from minus 1 to 10. Bars for most growth periods sit between 0 and 2 percent alpha. The widest bar appears at 7 years with an alpha of approximately 1.3 percent. A bar dips slightly to the left of zero at 13 years indicating negative alpha. Shorter bars at the top of the chart reflect smaller sample sizes for longer growth periods.

The best growth period was 7 years. Median alpha: 1.3% per annum. If the broader market grew 60% over 7 years, these recommendations grew 75%. That 15% extra does not survive exit costs. Capital gains tax, agent commission, and stamp duty on re-entry consume approximately 7% per trade plus CGT.

The worst period was 13 years. Alpha was negative by approximately 1%. Holding an infrastructure hotspot for 13 years produced less growth than the broader market.

What a Non-Infrastructure Algorithm Delivers

The SuburbData DSR (demand to supply ratio algorithm) does not consider infrastructure as a variable.

The charts below show DSR versions 1, 2, and 3 across the same growth periods. Every bar points to the right. No negative periods across any version.

The chart below shows DSR1 performance across all growth periods from 1 to 15 years. Compare the direction and length of every bar against the expert's chart above.

Horizontal bar chart titled DSR1 Top 20 Alphas 2010 to 2025. The vertical Y axis shows growth periods from 1 year at the bottom to 15 years at the top. The horizontal X axis shows alpha as percentage above the broader market from minus 1 to 10. The longest bar appears at 1 year with an alpha of approximately 5 percent. Bars shorten progressively toward the top as growth periods lengthen. All bars point to the right. No bars cross into negative territory across any growth period.

DSR1 delivered approximately 5% alpha in year one. Positive across every growth period tested. No infrastructure variable included.

DSR2 covers the same growth periods. Every single one points to the right. No negative periods. No infrastructure variable.

Horizontal bar chart titled DSR2 Top 20 Alphas 2010 to 2025. The vertical Y axis shows growth periods from 1 year at the bottom to 15 years at the top. The horizontal X axis shows alpha as percentage above the broader market from minus 1 to 10. The longest bar appears at 1 year with an alpha of approximately 7 percent. Bars remain consistently positive across all growth periods. The shortest bar appears at 13 years with an alpha slightly above 1.3 percent. No bars cross into negative territory.

DSR2 maintained positive alpha from 1 to 15 years. The worst period was 13 years at approximately 1.3%. It never went negative. No infrastructure variable included.

DSR3 shows the strongest performance of the three versions. The bars are longer across the short-term periods. Every bar still points to the right. The worst period still beats the infrastructure expert's best.

Horizontal bar chart titled DSR3 Top 20 Alphas 2010 to 2025. The vertical Y axis shows growth periods from 1 year at the bottom to 15 years at the top. The horizontal X axis shows alpha as percentage above the broader market from minus 1 to 10. The longest bar appears at 1 year with an alpha of approximately 9 percent. Bars decrease gradually as growth periods lengthen but remain positive across all periods. The shortest bar appears at 15 years with an alpha of approximately 1.3 percent. No bars cross into negative territory across any growth period.

DSR3 delivered over 9% alpha in year one. Worst period was 15 years at approximately 1.3% more growth than the national benchmark. No negative periods. No infrastructure variable included.

The Expert's Most Recent Work

Recent infrastructure-led reports from 2021 and 2022 showed improvement. 40 recommendations. Alphas above 4% across 1 to 4-year growth periods.

The charts below show those recent alphas alongside DSR3's performance for the same period across its top 10 local government areas.

The chart below shows the expert's recent infrastructure-led recommendations from 2021 to 2022. Four growth periods. All bars above 4%. This is the best the infrastructure method has produced.

Horizontal bar chart titled Infrastructure-Led Alphas for 2021-11 to 2022-11. The vertical Y axis shows growth periods of 1, 2, 3, and 4 years. The horizontal X axis shows alpha as percentage above the broader market from 0 to 15. All four bars point to the right. The longest bar appears at 1 year with an alpha of approximately 5 percent. The shortest bar appears at 3 years with an alpha of approximately 4.3 percent. All alphas sit above 4 percent across all four growth periods.

The chart below shows DSR3 for the same period across its top 10 local government areas. Same four growth periods. Same X axis scale. The difference is immediate.

Horizontal bar chart titled DSR3 Top 10 Local Gov Areas for 2021-11 to 2022-11. The vertical Y axis shows growth periods of 1, 2, 3, and 4 years. The horizontal X axis shows alpha as percentage above the broader market from 0 to 15. All four bars point to the right and extend well beyond the infrastructure expert chart for the same period. The longest bar appears at 2 years with an alpha of approximately 14 percent. The shortest bar appears at 4 years with an alpha of approximately 11 percent. All alphas sit above 11 percent across all four growth periods.

DSR3's worst period for that era was 4 years at approximately 11% alpha (growth of 17% pa, 87% total).

The expert's 4-year alpha was approximately 4.5% (growth of 10.5% pa, 57% total).

What This Actually Means

ABS data shows no meaningful correlation between infrastructure spending and capital growth. Where a correlation exists, it runs in reverse.
9 out of 10 individual projects produced no measurable uplift in surrounding property values.
An infrastructure expert with 500 recommendations over 18 years delivered a peak alpha that does not survive transaction costs.

Infrastructure research is not worthless. It is overrated.

The time spent on the next rail line or hospital announcement is time not spent on variables that actually predict growth.

Conclusion

Big projects. Little payoff.

Three ABS datasets. 120 specific projects. 500 expert recommendations. The finding is consistent.

Infrastructure spending does not reliably drive capital growth.

The markets that outperform are not the ones with the biggest projects. They are identified by more reliable variables that measure supply and demand.

EBS 14 Infrastructure Projects: Big Projects, Little Payoff