But how do you balance those trade-offs? It’s common for cities to quantify when a freeway might hit the point of congestion, or when a sewerage system might overflow, but it’s much less common to measure how much damage is done by not allowing people to live where they want to live.

This post outlines a novel methodology I constructed with the help of the team at YIMBY Melbourne (including superstars Paul Spasojevic and Jonathan O’Brien) to measure the barriers created by different levels of restrictive zoning in cities.

While there’s strong evidence that zoning limits the overall supply of housing, you don’t need to believe this fact for our methodology to be useful. Even supply sceptics agree that planning changes where housing is built in a city – often away from where people most prefer and towards where planners think is better in the long run.

The distortions created by zoning may well be justified, but to know for sure, planners should be required to measure these distortions as part of any zoning or strategic plan. Where the distortions are greatest, more care should be taken to ensure the costs of restrictive zoning are worth the benefits.

The downside of zoning

Most westerners live in detached homes. In the absence of restrictive zoning, every home sale would be a bidding war between a potential single homeowner and people who each want to pay less to share the land by building apartments or townhouses.

Where land prices are low, the single homeowner will likely win out. Constructing new apartments is expensive, and if the price people are willing to pay to live in apartments in the area is low then the homeowner is likely to win at auction.

But as a city becomes more popular, the number of people and the price they’ll pay to live in apartments increases and so those apartment buyers bid up the price of the land until they can outbid prospective homebuyers and start building.

In areas with restrictive zoning, the price of land doesn’t respond to demand from apartment dwellers. Apartment prices in an amenity-rich area might be so high that you could justify a 100-storey building, but if it is illegal to build then the price of land will only reflect what somebody is willing to spend to use that land as a single-family home.

The gap between what people would be willing to pay to put apartments on a block of land and the current price of that land is an effective measure of how much zoning is restricting housing on a given lot.

Planners should measure the distortions created by zoning in their city

With the right data, measuring the costs of restrictive zoning in cities is relatively easy. For a given plot of land, all you need to do is estimate what the land is currently worth, what an apartment on that land would sell for, and the cost of building apartments at various building heights. If enough density is permitted, the price of land should reflect these calculations. Where there are super-profits to be had by building on a block of land and yet those profits haven’t either jacked up land prices or lowered apartment prices, then zoning is restricting supply on that lot.

Unlike other measures of distortion created by zoning, this method allows us to go beyond comparing the costs of current zoning to a fictional world without any zoning at all. Planners can also compare how current zoning compares to other politically achievable zoning outcomes.

YIMBY Melbourne wanted to see how much Melbourne’s current zoning restricts supply compared to 6-storey height limits being rolled out in Sydney. When we did that in one of Melbourne’s richest areas (Boroondara) the results were stark. If a developer could build 6 storeys in this area they could make on average $160,000 per apartment built, even after their usual profit goal of 20%. This kind of extreme profitability should either make apartments cheaper or be priced into the land. The fact that neither has happened shows that zoning is restricting supply.

The super-profits available suggest that zoning in Boroondara is incredibly restrictive. But the same zoning might not be restrictive where demand is lower. In Brimbank for instance, where there is much less demand to live in apartments, zoning to 6 storeys would not change development much. People who want to live in a house in Brimbank would out-bid apartment dwellers, even if we allow 6-storey tall buildings. As a result, we can say that the existing zoning does not significantly reduce the supply of new housing on existing lots in this area.

We’ve calculated the distortions created by zoning for inner and middle Melbourne and the situation is dire. Zoning in areas like Boroondara is distorting the value of land, to the detriment of people who want to live in apartments. 

Planners should quantify the costs of restrictive zoning regularly and publish their findings. Measures of distortion created by zoning in each local area should be a metric that keeps planners up at night. This figure should be calculated for each neighbourhood every quarter, and when the number grows and yet planning applications do not then the warning lights should flash red. 

The costs of zoning should drive housing targets

When setting housing targets for a city, the distortions created by zoning should be the most important variable in deciding where homes should go. In Melbourne, we allocated the city-wide housing target to Local Government Areas based on the total number of super-profitable apartments that would be possible in each area if 6 storeys were allowed near train stations and tram stops.

The targets ensure that when the state upzones an area, local governments are prevented from using other planning measures to slow construction. The targets are higher in areas like Boroondara where there is ample public transport, lot sizes are large and demand for apartments is high.

In expensive areas, we opted to recommend less up-zoning in lots that were far away from public transport, because the cost of building more extensive transport infrastructure in those areas risks being greater than the harm created by restrictive zoning. But this tradeoff was merely a ‘rule of thumb’ for YIMBY Melbourne. In Victoria, the state government recently merged the Department of Planning with the Department of Transport so that these kinds of trade-offs could be better measured.

“Zoned capacity” is a terrible way to measure the effects of zoning on a city

There's another method that is popular amongst planners to decide whether zoning is affecting supply. Planners will count up the theoretical maximum number of homes that can be built using existing zoning (zoned capacity), and if that figure is greater than an arbitrary housing target for the next few decades, they will declare that zoning is not restrictive.

This method was popularised in Melbourne by Michael Buxton, Joe Hurley, and Kath Phelan. We can see this idea used in Boroondara’s housing strategy:

However a zoned capacity greater than the need for housing does not prove that current zoning is appropriate. Even if zoned capacity exceeds predicted demand for housing it can be restrictive for several reasons:

1. Some land will never be developed and so the ‘zoned capacity’ provided by it is not real

The existing homes on a block of land have value. There are some homes in Boroondara where 3-storey apartment blocks are permitted, but our analysis suggests that once you pay for the land, demolish the house, and build something new there isn’t enough profit to make it worthwhile.

The homes in the graph above are from just a single zone in Boroondara but represent 12,000 homes of Boroondara’s zoned capacity. Almost all of it is unbuildable. The only way to make 3-storey buildings on this land profitable would be to make housing so unaffordable that the price of apartments increases significantly, harming renters who rely on apartments to stay housed.

Even if there are technically just enough “profitable” apartments to meet demand, zoning still restricts supply. That’s because not every block of land is suitable for new housing. Some people love their homes and will refuse to sell. Some existing non-residential uses for land might be very profitable or beloved by the community, making development unlikely. Additionally, council heritage and neighbourhood character guidelines may also act to limit what can be built on a given block of land.

Each of these factors chips away at the real zoned capacity of a given area. For zoned capacity to actually enable enough housing to be built, the number of potentially profitable apartments in an area must be many times the number of homes a community needs.

2. Housing ‘need’ numbers are often made up.

Boroondara estimates that 9,400 more homes will be needed in their local government area over the next 15 years, but these housing needs figures are often made up. In Victoria’s case,  the population estimates for each area usually come from the state government's Victoria in the Future projections, which are designed to predict how many people will move into an area.

But these projections are based on each given area's “capacity to absorb extra population” – meaning that zoned capacity has dictated future population projections, which are now being used to justify existing zoned capacity.

The real number of apartments that Boroondara needs is however many will make rents cheap. But zoned capacity does not provide any insight into what that number might be.

3. Rationing zoned capacity gives a monopoly to developers and encourages land banking.

Despite the limitations outlined above, the current zoned capacity of 65,000 in Boroondara means there will be some suitable lots to build on. But because the need for housing and the zoned capacity are so close to each other, the people who own the best lots for development are able to maintain a near-monopoly on new housing in the local area.

If you hold one of the only possible lots that apartments can be built on in a local area, it is much more tempting to hold off on building in the hope that prices will be higher in the future and your profits will be greater.

Where upzoning is broad, land speculators have to look over their shoulders and consider the many more nearby land-owners who might beat them to the punch, discouraging speculation and lowering the cost of new apartments.

In the example of Boroondara above, land prices have been suppressed by $160,000 per potential apartment because of highly restrictive zoning. Narrow upzoning would cause property prices to rise by around this much, as land speculators know that there are future residents in the area who will have no choice but to buy an apartment on their land. If there is broad upzoning then apartment buyers will be able to play landowners off against each other, and as a result they will get much more of that $160,000 in the form of cheaper house prices and lower rents. 

4. Only allowing a small amount of “zoned capacity” denies people choice

When planners limit zoned capacity to a number that’s only slightly higher than what needs to be built, it denies people choice. Some people might prefer to live on a side street, or they might prefer to live in a particular part of a local area. Limiting zoned capacity means they are denied that choice and will end up with housing locations that don’t suit their needs as well.

Planners attempt to account for this by recommending a diversity of housing options in their strategic plans – but it’s impossible for them to perfectly predict what kind of housing people will want and where. Increasing zoned capacity to many times future demand will maximise choice, and reduce the number of tradeoffs people have to make when choosing where to live and work.

For all of these reasons, big changes in zoned capacity are needed to shift the needle on housing supply. The exact amount of capacity needed is unknown and will depend on what type of zoning capacity is provided, building costs, land values, and demand for apartments in a given city.

But Auckland does provide us with some clues on how changes to zoned capacity affect prices. In 2015 Auckland tripled its zoned capacity, and over the next 6 years the number of homes in Auckland increased by 4.11% more than would have otherwise, lowering rents by 14-35%.

YIMBY Melbourne is proposing that the inner and middle suburbs of Melbourne be upzoned such that their zoned capacity is increased 7-fold. Many of our upzoned areas include the most desirable land in Melbourne, increasing the chance that the zoned capacity will be used to build new homes.

Giving up control

One of the hardest things about broad upzoning is that it requires planners to give up control.

Some of the most beautiful cities in the world have streets with rows of buildings all of a similar height. These symmetrical streetscapes are largely the result of technological limitations that made tall buildings difficult. For instance in Brooklyn, brownstones are usually a similar height because pumping water to higher levels was expensive. In Europe, fire regulations and a lack of elevators drove many old cities’ building heights.

Planners often dictate the maximum heights of buildings to achieve similar aesthetic outcomes, even when they do allow tall buildings to be built. For instance, the new Melbourne Suburban Rail Loop master plans have a complex set of zones designed to achieve aesthetic outcomes unrelated to the area’s overall density:

Planners no doubt hope that demand for housing is so strong and zoned capacity so limited that apartment buildings will be built up to somewhere approaching the maximum zoned capacity of each zone. As a result, their preferred streetscape aesthetic will be achieved.

But broad-based upzoning means giving up control. If instead of allowing one street near Box Hill Station to have 20 storeys we decide to allow 6 storeys along the whole train line, planners will not be able to decide where new apartments go. No doubt we will end up with streets where half the homes are detached and the other half have apartment buildings of various heights. The process will not be ordered, but it will mean that there are many more potential sites available for people who want to build apartment buildings, so the cost of upzoned land will be less.

Giving up some aesthetic control means creating a city much more like Tokyo where building heights differ markedly even on lots next door to each other.  In these cities, planners need to constantly measure demand and supply changes, and design the delivery of infrastructure in response to these measurements—rather than trying to control demand and supply itself. 

While there are real costs to more permissive zoning (e.g. local traffic congestion), these costs should be measured and balanced against the distortion created in housing markets by the zoning status quo. This is a marked improvement on current planning regimes, which mostly neglect to measure the costs of decisions at all.  

By broadly upzoning and enabling more choice, cities everywhere will empower more people to live where they want to live.

Maintaining clean air in the bedroom is more difficult than it sounds. When you exhale you breathe out carbon dioxide. Even with a HEPA filter CO2 can build up, reducing the quality of your sleep and making it more difficult to focus the next day.

The benefits of natural light in the morning are far less well established. Many people find that a sunrise helps them wake up, but there’s no evidence natural light is any better than a sun-rise lamp paired with plenty of bright natural light throughout the day.

Opening windows can release CO2, and planning regulations for apartments often require windows in bedrooms to let fresh air in. But most apartments in cities are on noisy and polluted streets. It’s very difficult to maintain a quiet and dark environment while keeping the bedroom windows open to let in fresh air. And as air conditioning becomes more common, opening windows also results in losing precious cool air.

Good urban planning can reduce street noise, but even open windows on quiet streets let a city’s light pollution in. We need to accept that most apartment dwellers are better off leaving their blinds down with their windows closed at night.

Closing windows ensures the bedroom stays cool, dark, and quiet, but how do you get fresh air in? The best solution is mechanical ventilation. Heat transfer units (aka energy recovery ventilators) swap dirty inside air for filtered outside air, recovering the heat along the way so that the inside stays the right temperature.

Once the purview of wealthy eccentrics on Grand Designs, these machines are now reasonably affordable if they’re installed when an apartment is built. Mechanical ventilation is the only real way to ensure a good night’s sleep in a small bedroom on a busy street.

In a world of housing abundance, nobody would build apartments without bedroom windows because they would never sell. But given zoning restrictions artificially raise rents, allowing well-ventilated window-less bedrooms is one way to even out the score.

Subscribe now

Why are windows in bedrooms good?

Public health was the main reason that window mandates were passed in many Western countries. Wooden apartments without sprinklers, polluting gas stoves, and tuberculosis risk made the public health case for windows inescapable.

But things are not as bad as they were in the tenement era. We now have electric stoves and HEPA filters that can filter everything except CO2, making indoor air quality better than even outdoor air. And deadly fires are much less common than they once were, with sprinklers largely mitigating the risk of not having easy egress through a bedroom.

What’s left now are two main benefits to windows in modem apartments:

• They can be opened to reduce levels of CO2, improving sleep quality; and

• They provide natural light in the morning that may improve your circadian rhythm.

Both of these are great benefits to having a window in the bedroom, but the health advantages they provide are not as big as you might think.

Clean air in the bedroom can improve sleep quality and health

High CO2 can make a room feel stuffy, and it has been shown to reduce the quality of sleep - even in the absence of other pollutants.

There have been numerous studies that have assessed sleep quality at varying CO2 levels. They usually involve a small number of participants. Some have been purely observational while others have involved fans that pump out CO2 when it is too high. One study even randomized participants to open v close windows overnight.

The best estimate I have from one Australian study is that for every 100ppm increase in CO2 concentration, deep sleep falls by 4%. Sleep quality is unaffected at CO2 levels below 700, steadily rising from 700 and above. At levels higher than 2,600, cognitive performance has been shown to decrease the next day.

I have a CO2 monitor in my bedroom, and if I leave the doors and windows closed CO2 rises to levels that will definitely affect my sleep:

Mechanical ventilation is better than windows because people need to close their windows overnight

Windows might be great for reducing CO2, but they won’t help if you leave them closed.

In the anglosphere’s post-war suburban ideal, most people lived in a moderate climate on quiet streets with leaky houses and cross-ventilated windows that could stay open all night even during the warmer months. But for new apartment dwellers, energy efficiency standards require homes to be airtight. Cool air also improves sleep, and people rely on air-conditioning to maintain an even temperature. Opening the bedroom window means letting the cool air escape and exposing yourself to noisy and polluting traffic outside.

It’s for this reason that I doubt regulations mandating windows will change indoor CO2 quality. There should instead be much wider installation of energy recovery ventilators (ERVs) units during building construction. These devices have fans that swap dirty indoor air with filtered outdoor air while keeping the nice temperature that the indoor air already has.

ERVs bring in fresh air from outside, in contrast to heat pumps or reverse cycle air conditioners that tend to re-circulate existing air. ERVs are difficult to retro-fit in existing apartments, making it essential that they are considered when new apartment buildings are being designed.

Light improves circadian rhythm

Natural light is clearly good for human health, it’s just not clear how much having a natural sunrise improves your day.

Light when you’re trying to get to sleep is bad. If you put people in a lab and give them bright light in the evening their melatonin will be suppressed, which makes it much harder to sleep at the right time. Even light from 1-5am increases cortisol levels, presumably reducing sleep quality. Real-world studies have found that people who sleep in bright bedrooms at night toss and turn in bed and report poorer quality sleep.

But it’s not ideal to live in pitch black all the time. If you deprive people of all light, they tend to go to bed later and later each night, until their sleep cycle is completely reversed. Daylight suppresses your melatonin, helping you maintain a 24-hour cycle.

Morning light also seems to help. When researchers woke up lab participants at 6am with bright lights they got up early again the next morning even without the stimulus. And when people stuck in an Antarctic winter with little daylight were exposed to bright light in the morning it shifted their sleep cycle, helping them to rise and improving their cognitive function. But in both studies the morning light didn’t help people get to bed earlier, so total sleep time was reduced - at least in the short term.

But natural light from bedroom windows probably isn’t that important

So how much light do you need in the morning? And at what time?

This study found that even a bedside dawn simulator can be effective at improving self-assessed sleep quality. Another study found that 10 lux (a bright light bulb) was enough to improve your circadian rhythm in a lab setting.

It’s also not at all clear that you need bright light before you wake up. Getting bright light only after waking was enough to raise cortisol levels in this study, meaning that getting out of bed and enjoying some living room sun could have real benefits. Office workers are more depressed if they don’t have natural light throughout the day, and so the Munger dream of a completely dark living environment is probably not a good idea.

All of these studies are very short-term. What you won’t find above is evidence that natural light in the bedroom is good for your long-term health. If you personally find a sunrise makes it easier for you to get out of bed then it’s worth it for you. If you like to keep the blinds closed that’s fine too.

Should we mandate windows in bedrooms?

Windows in bedrooms do have some benefits, but they also have real costs. If you require windows, fewer apartments can be built on a site. When that site is developed, fewer people will get homes. There will be a longer queue at open for inspections, and rents will rise.

Regulating for windows in bedrooms can be especially troublesome because some windowless bedrooms still let in lots of light. “Borrowed light” is where windows allow light through a living room to the outside. They provide enough light so that you know what time of day it is, but not a bright sunrise.

This kind of set up can work really well for one-bedroom apartments where privacy between the living room and the bedroom doesn’t really matter.

Requiring good ventilation in bedrooms on the other hand will affect rents much less. Because these regulations increase construction costs without decreasing the amount of housing that can be built, the effect on housing supply for in-demand areas will be much less, and more of the cost will be paid for by owners of land.

What this means for you

For most existing apartments, mechanical ventilation is currently too difficult to install. And the technology is not quite there yet to provide cheap window-mounted ventilation units for renters. So what can you do to improve your sleep?

Leaving your window open all night might still be the best option, but you might also consider briefly opening the window before bed to lower CO2 levels before you sleep (The Germans call this “Lüften”). If you live alone you can also leave the door to the living room open to expel some of your overnight CO2 into that larger space.

Try your best to make the room quiet and pitch-black at night. If you can time your blinds to open around sunrise that’s a bonus, and if not a bright sunrise lamp will probably have a similar effect. In summer, keeping the bedroom cool can also help to improve sleep quality.

Home CO2 monitors are becoming much more common, so if you have any overnight readings post them below with details of your room!

One of the reasons I made the website is because locals often overstate the density of their area when they oppose new homes.

Unfortunately we’ve had a massive case of this error in Melbourne’s “The Age” yesterday, where the paper has claimed that parts of the city are “more crowded than Manhattan”. Of course this is not true, but how does the mistake get made, and why does it matter?

No, Melbourne is not more crowded than Manhattan

This might seem so obvious it’s not worth saying, but here are the stats regardless. The Global Human Settlement Layer (GHSL) divides up Manhattan into 1km squares, which we can then compare to 1km squares around Melbourne.

Even Melbourne’s most dense area (The North East of the CBD) is much less dense than Manhattan’s busiest neighborhoods.

If you compare Melbourne to Paris, the lack of density is even more stark:

Southbank, the CBD and Docklands are reasonably dense compared to the financial district, but as soon as you look more than 1km out, NYC is 3-4x more dense than Melbourne.

Why do people measure density so badly?

The challenge here is to compare like with like. The Age calculated the density of Southbank which is a tiny suburb not including the massive park right next door. They've then compared that to the entire island of Manhattan, including central park.

You simply can’t compare the density of differently sized areas. That’s because small areas often have wildly variable population densities. One area might be full of housing, while nearby areas have parks, schools, offices, or stadiums.

As an example, shown below is a small area of Darwin that's more dense than the whole island of Manhattan. It's a meaningless comparison.

The difference is obvious when you take it to this extreme, but the problem exists whenever you compare the density of areas that aren't exactly the same size. Hence the need for the km square grid. Citydensity solves the problem by only comparing the density of tiles on the grid, so that you are always comparing like with like.

What does this mean for Melbourne’s growing suburbs?

This might feel like graph-nerd ephemera but the policy implications are huge. The data is said to “explode the myth that Melbourne’s inner and middle ring has not been growing to accommodate more people.” But instead of exploding a myth the article has craeted a new one, confusing people into thinking that Melbourne is already very dense.

In reality the biggest density difference between Melbourne and cities like London or Paris is in areas like Hawthorn and Fitzroy North that just aren't very dense. Melbourne really falls behind in the middle ring, 3-20km away from the city.

And things are not getting much better. With the exception of City of Melbourne, all of our inner and middle Local Government areas have been growing at a rate slower than the Melbourne average, leaving much of the heavy lifting to greenfield suburbs where infrastructure is poor.

That’s not to say that no new housing is being built in inner Melbourne. For the past decade the city has been following Matthew Guy's idea to relegate density to the most polluted, noisy, and dangerous streets. This works well for developers because it’s easier for them to corner the market, and wealthy homeowners are also happy since there is little development on their leafy green side-streets. But it’s inequitable and it's making the city a bad place to live.

The worst example of this is City Road in Southbank, a 4 lane death trap straight through one of Melbourne’s most dense areas. It’s great that the people who live in these new towers have a place to live but their lives would be a lot better if we’d allowed density on nicer streets as well.

There is an easy solution. We need to fill the gap in our missing middle by allowing 6 story human scale developments on every lot within 500m of a train station or tram stop. It won’t just make rents cheaper, it will also make Melbourne a much more enjoyable place to live.

YIMBY Melbourne has a great plan that goes through all the details.

The way we define what a ‘city’ differs across countries and cultures. The density of different areas is dependent on essentially random political boundaries. So Australia tends to have huge cities, whereas there are over a dozen cities near NYC that would be considered suburbs in Australia.

It’s impossible to compare the density two cities if we are not comparing apples to apples.

But it’s not just across cities. When we are deciding how dense the upper west side is in Manhattan, should we include central park or not? That choice could halve or double the density of the neighbourhood - but there’s no system in place to do so fairly.

Luckily there is a solution. Today I’m launching CityDensity.com to compare the world’s big cities.

The website works by breaking the world up into tiny squares, with each square having 1km sides. Because the squares are essentially random they’re also fair - no political borders can make a city seem bigger or smaller than it is.

The two best measures of city density are

• Population weighted density which measures what it feels like to live in a city - how tall the buildings are and how many people are likely to be on the street.

• Population which measures how many people live close by to a city

Compare what it’s like to live in or visit a city - population weighted density

To understand what it feels like to live near or visit a city, look at “population weighted density”. This is a measure of the typical density experienced by a person who lives in a given city. This measure takes account the fact that some cities feel more dense than official statistics show. Their density might be lower because of oceans, mountain ranges, or zoning that concentrate populations in small areas. One example of this is Hong Kong. It is one of the most dense cities in the world but raw density statistics are much lower because of all the mountain ranges and oceans.

Population weighted density represents the typical density experienced by somebody who lives in a given area. It measures how dense the typical neighbourhood is. It is calculated by finding the average population density of each 1km square in a region, weighted towards the squares with a greater population.

Compare the economic pull of a city - population

Cities are more than just a collection of neighbourhoods. When lots of dense areas are grouped together they create dynamism much more than the sum of their parts. While Hong Kong is incredibly dense, Tokyo is dense in every direction and so millions of more people call the city home. To understand the economic power of a city it can be helpful to look at the total number of people who live nearby, without worrying about things like water or mountain ranges.

City Densities lets you select many cities and compare, but if you want to look at a specific point anywhere in the world try out Tom Forth’s amazing website that uses the same data I use here.

Other density measures

I’ve included lots of other density measures on the website. Most of the options show who lives in rings around a given city, while “cumulative” measures consider who lives within the whole area from the center of the city up until that point.

How good is the data?

The team at Global Human Settlement Layer made the map of tiny squares, with data collated from around the world by NASA. Most of the populations are projected forward to 2020 from national censuses. Many of the original data is from around 2010, and so is starting to lose accuracy in fast growing areas. NASA are updating their data in Spring of 2024, and hopefully GHSL will not be far behind. For Australia I’ve overcome this by using the ABS’s 2022 km grid, which is far more accurate for Melbourne and Sydney’s fast growing outer suburbs.

For each road in Melbourne, and many in Bendigo, Geelong, and Perth, you can see each individual passing event that was collected by our 175 cyclist volunteers. Data from the map has been invaluable to local councils planning infrastructure ugprades in their area.

Try out the map yourself by visiting www.passbox.org/map

This study describes the design, key findings, and process for policy translation of a series of modelling studies conducted for the Victorian Department of Health COVID-19 response team during this period.

Scott N, Abeysuriya RG, Delport D, et al. COVID-19 epidemic modelling for policy decision support in Victoria, Australia 2020–2021. BMC Public Health. 2023;23(1):988. doi:10.1186/s12889-023-15936-w

This was the largest study to-date to record passing behavior on public roads, and a large number of road and traffic attributes that might affect passing distance were included in the analysis.

We found that even on narrow roads, putting in a bicycle lane helps. When drivers pass cyclists on roads with painted bicycle lanes, they tend to give more space than on roads without bicycle lanes. This is true even when controlling for the space available on the roadway. Drivers also travel in a more predictable fashion, with less variability in passing distances when a bicycle lane is present.

Protected bicycle lanes completely remove the risk of passing events less than 1 m. However, where it is not possible to build a protected bicycle lane it is preferable to have a painted bicycle lane than no bicycle lane at all. Other protective factors include: wider lanes, single lane roads, smaller vehicles, and the removal of on-street parking.

The full study is available here: Nolan J, Sinclair J, Savage J. Are bicycle lanes effective? The relationship between passing distance and road characteristics. Accid Anal Prev. 2021;159:106184.

The rules required those who had previously been paid more than JobKeeper to stay on their old wage, with the rest topped up by their employer. Those who had previously been paid less (one quarter of them) got a pay rise.

It’ll be less rewarding from the end of September. That’s when it’ll fall to $1,200 per fortnight for people who had previously been working 20 hours or more hours per week, and $750 per fortnight for people who had previously been working less than that.

At the same time, the temporary coronavirus supplement paid to Australians on JobSeeker and other benefits will fall from $550 a fortnight to just $250.

Those receiving only the full-time JobKeeper rate can expect their incomes to fall by 20%. Those on the part-time rate can expect their income to halve.

But the good news is that at least some will be able to top up their incomes by applying for JobSeeker.

JobKeeper and JobSeeker

Many will be able to apply to recieve both.

They will need to satisfy the assets test for JobSeeker, which is being re-imposed from September 25. It will deny JobSeeker to single homeowners with assets of more than $268,000 in addition to their home, and to single non-home owners with assets of more than $482,500.

They would also need to wait for a reimposed liquid assets waiting period of between one and 13 weeks, depending how much money they have in their bank accounts.

Their partners would need to earn less than $3,068 a fortnight, or roughly $80,000 a year.

And they will be required to make at least one phone or online appointment per week with an employment services provider.

Up to $554 on top of JobKeeper

Our calculations suggest that under the new rules from late September, such a person on the part-time JobKeeper rate of $750 a fortnight should be able to claim up to an extra $554 in JobSeeker – taking their total income to $1,304 per fortnight – only $196 per fortnight less than they got when JobKeeper was $1,500 per fortnight.

It gets better. Even people getting the new lower full-time JobKeeper rate of $1,200 per fortnight will be able to get some JobSeeker if they fit through the hoops.

Our calculations suggest they will be eligible for up to $284 per fortnight JobSeeker top-up, taking their total income to $1,484 for fortnight, only $16 per fortnight less than they are receiving now.

It gets better still. If they pass through the hoops, they will also become eligible for other benefits such as a Commonwealth Health Care Card, Family Tax Benefit part A if they have kids, and rent assistance if renting.

The treasury believes 245,000 Australians will be on both JobKeeper and JobSeeker by the end of the year.

You’ll have to apply

One of the virtues of the original JobKeeper was that, from the point of view of the recipient, it was automatic. Once their employer decided to apply for it, there was nothing else they needed to do.

As JobKeeper is phased down and JobSeeker returns to its traditional role of supporting Australians on low incomes, there will be a lot more they need to do.

Some won’t bother, and some won’t succeed, but at least until the end of the year it’ll be possible for some Australians on JobKeeper to get more or less what they were getting before.

It’s less good for those pushed out of work

It’s worth sparing a thought for those that will lose their jobs as JobKeeper winds down.

In October, employers wanting to stay on JobKeeper will have to be retested and approved, and in January retested and approved again.

Many will miss out. Retesting is expected to reduce the number of workers on JobKeeper by 60% over the last three months of this year and by a further ten percentage points over the first three months of next year.

If those whose lose JobKeeper also lose their jobs — and many will — they’ll have to make do with the much lower JobSeeker payment of about $825 a fortnight, not much more than half of the $1,500 a fortnight they had.

Treasury expects 1.5 million people to be on JobSeeker by the end of the year.

It expects some 245,000 to recieve both JobSeeker and JobKeeper. That will leave around 1.25 million to get by on the lower JobSeeker alone.

by Brendan Coates, Jonathan Nolan

Published at The Conversation, 22 July 2020

Homelessness is highest in Northern Territory, more than 10 times worse than any other state and territory. The rate of homelessness has risen across most states, and is up 10 per cent nationwide between 2006 and 2016. The largest increase came from a larger number of people living in severely crowded dwellings.

There has been much attention on the emerging homelessness crisis among the elderly. But most homeless Australians are young: about 58 per cent are younger than 35 (Figure 2.2). About 81 in every 10,000 Australians aged 15-24 are homeless, falling to 39 in every 10,000 among those aged 55-64. In fact people aged 19-34 account for 65 per cent of the increase in homelessness in Australia over the five years to 2016, and men account for 69 per cent of the increase.

In contrast, relatively few older women are homeless. But the risk is rising: nationally, rates of homelessness increased fastest over the five years to 2016 among single women entering retirement age. Declining rates of home ownership among younger poorer Australians, together with the lower retirement savings of many women, increase the risks that more older women will suffer poverty and homelessness as they approach retirement.

How do people become homeless?

Homelessness is not an individual problem. A person who becomes homeless is likely to have faced a difficult combination of economic and social circumstances, but a key ingredient is almost always expensive rent. Australian evidence shows that the risk of a disadvantaged person becoming homeless increases from 2 per cent in areas with plentiful cheap housing, to 17 per cent in areas with high rents.

And rents are getting more expensive. While private renters in Australia are spending a similar share of their income on rent, the disappearance of social housing means that overall, the bottom are spending more.

The social security safety net has fallen out from under us

Income support system plays an important role in alleviating financial stress and poverty for low-income Australians. Yet while the Age Pension, Parenting Payment, Carer Payment, and Disability Support Pension are indexed to wages, JobSeeker, Youth Allowance, and rent assistance only increase with inflation. These payments have therefore become woefully inadequate as a safety net for unemployed Australians. Unlike wages or pensions, JobSeeker has not increased in real terms in more than 20 years.

This has ‘squeezed’ the living standards of people living on JobSeeker relative to the rest of the population. Households of working age receiving JobSeeker are under much more financial stress than households receiving other welfare payments.

At the same, more people are having to rely on JobSeeker who would have been better supported by the social welfare system in the past. Eligibility for the Parenting Payment and the Disability Support Pension has been tightened in recent years. Changes in assessment rules for the Disability Support Pension in 2012 have especially hit Australians approaching retirement age who have musculoskeletal health problems. The changes coincide with an increase in the number of older people on JobSeeker.

Social housing can also help

The public health crisis caused by COVID-19 has resulted in an unplanned trial of a ‘housing first’ approach. Many people sleeping rough who would have previously been denied access to safe housing were given hotel vouchers, to help them socially isolate through the crisis. By one estimate rough sleeping was reduced by 60 per cent almost overnight.

But people rough sleeping cannot stay in hotels forever. Despite its importance to vulnerable Australians, Australian governments have invested little in social housing over the past two decades. The number of social housing dwellings has barely grown over that period, while Australia’s population has increased by 33 per cent.

Tenants generally take a long time to leave social housing: most have stayed for more than five years. And when people do leave public housing, it is rarely by their own choice. In 2012, only 18 per cent of people leaving social housing voluntarily entered the private rental market or bought their own home. As a consequence, there is little ‘flow’ of social housing available for people whose lives take a big turn for the worse, and many people who are in greatest need are not assisted. The result is that fewer Australians are living in social housing than in the past, and every year proportionally fewer social housing units become available for new tenants.

How to tackle Homelessness in Australia

There is a powerful case for more Commonwealth Government support to reduce homelessness and help house vulnerable Australians. But housing subsidies are expensive and not all policies are equally effective. The Government should prioritise reforms that target those most in need, and deliver the biggest bang for taxpayers’ buck.

The Government should give priority to constructing new social housing for people at serious risk of homelessness, replicating the success of ‘housing first’ programs abroad. Social housing is particularly effective stimulus, but does come at a cost. Once more units are constructed, they should be reserved for those most in need, and at significant risk of becoming homeless for the long term.

Boosting Commonwealth Rent Assistance by 40 per cent is a cost-effective way to help the much larger number of lower-income earners struggling with housing costs.

Housing will become substantially more affordable for most low-income Australians only if we build more of it. The Commonwealth Government should provide incentives to the stats to fix planning rules that prevent more homes being built in inner and middle-ring suburbs of the capital cities. More private housing can and does help those on low incomes by lowering the rents they pay. And lower rents reduce the risk of homelessness for those already vulnerable.

Housing vulnerable Australians poses substantial policy challenges. Past governments have refused to face up to the size of the problem because they feared doing so would fuel demands for massive new expenditures on housing. But the challenge is not insurmountable. Policy can make a difference, but only if we make the right choices.

Adapted from Coates B, Nolan J, Chen T. Tackling Homelessness in Australia: Submission to the Standing Committee on Social Policy and Legal Affairs’ Inquiry into Homelessness in Australia. Grattan Institute; 2020.

Australia is not far behind New Zealand in daily coronavirus cases, and we are catching up. We are currently recording about 100 cases a day, which is only slightly more than New Zealand’s peak of 87 a week ago. Most individual states in Australia are already ahead of New Zealand, and if they stay the course and keep Stage Three restrictions going they could beat New Zealand to 0.

Growing epidemiological evidence points to the possibility of eliminating coronavirus. Recent modelling by the University of Sydney found that with spatial distancing the virus could be all but gone from Australia by the second week of June, and improved contact tracing could probably prevent a resurgence of the disease.

Separate modelling published on Friday by researchers at the University of Melbourne reaches the same conclusion. And New Zealand modelling previewed by Jacinda Arden on Friday suggests that country has a better than even money chance of being COVID-19 free by late June, just two-months away.

In contrast, early modelling released by the Doherty Institute for the Australian Government was designed to model healthcare capacity and did not consider scenarios with the degree of spatial distancing we are achieving in Australia today. While it was helpful in preventing a catastrophe, It does not help us understand if eliminating COVID-19 is possible in Australia, and by when.

The prospect of 90 days of Stage Three restrictions is daunting, but there are good reasons to believe we could bring cases down to zero earlier than that. The University of Sydney and the University of Melbourne models simulate enough spatial distancing to halve the number of coronavirus cases roughly every 16 days. But for the past week Australian cases have been halving every 4.2 days. Even if that trend slows over the next few weeks, there’s a real prospect that many states could begin easing spatial distancing as early as May, with a clear plan in place by Anzac Day.

Getting from ‘nearly 0 cases’ to 0 cases is the toughest part of the elimination strategy. We would need to trace down every single case, isolating and testing all the people they might have come into contact with. Some towns or suburbs might need to stay on Stage Three restrictions for longer to stamp out the last remaining cases, and some schools might have to be closed again if cases flare up. We would need to quarantine international arrivals until a coronavirus vaccine is found.

But a successful elimination strategy would give most Australians a good chance of resuming a near-normal life well before the end of the year. If we succeed, crowds could return to the MCG in time for the Boxing Day Test.

Adapted from Duckett, S, Nolan J. Australia should join New Zealand and shoot for eliminating coronavirus. Grattan Institute; 2020

Early withdrawals may not mean big falls in retirement incomes

Our modelling suggests that a 35-year-old who takes the full $20,000 allowed from their super will see their balance at retirement fall by around $58,000.² For someone earning the median wage of about $60,000 today, they can expect their total super income through retirement to fall by around $80,000 in today’s dollars. But their total retirement income would fall by only $20,000 in today’s dollars, or around $800 each year.³ The Age Pension is means tested – the higher your super balance, the less pension you get. So workers who take money from their super will lower their super balance at retirement, but they will receive more Age Pension – helping to soften the blow.

The story is a little different for high-income earners. The total retirement income of a worker earning around $110,000 a year today (more than 90 per of workers of that age) would fall by around $70,000, or around $2,700 a year, if they withdrew the full $20,000 from their super over the next few months. They would lose the same amount in superannuation as a middle-income earner, but receive less Age Pension to compensate, since they’re ineligible to receive an Age Pension for most or all of their retirement.

Of course, many workers would be withdrawing their super at a bad time: the value of the ASX 200 has already fallen by about 40 per cent in the past month. If markets recover next year, that $20,000 in withdrawn super could soon enough be worth about $30,000 again. If that’s the case then the hit to retirement incomes is larger: $33,000 over the course of retirement for the median worker and $106,000 for a worker at the 90th percentile of all wage earners. ⁴  But for many Australians facing bankruptcy in the coming months, it’s a trade-off that may well still be worth it.

Putting cash in people’s pockets now is the priority

Australians are in the middle of an enormous liquidity crunch due to COVID-19. More businesses are closing by the hour. Casual workers will be particularly hard hit, but many part-time and full-time employees will soon become unemployed or have their hours reduced.

Most workers, including many on middle and high incomes, don’t have much money in the bank for a rainy day. Half of Australian working households – those with at least one employed person – have less than $7,000 in the bank, including mortgage offset accounts. About 20 per cent have less than $500.

The Government is temporarily doubling the rate of Newstart and accelerating people’s access to the payment by relaxing the assets test. But even a higher rate of Newstart won’t cover many people’s costs, especially people who are on higher incomes. Even at the top, about 40 per cent of the highest fifth of income earners have less than 4 weeks’ income in the bank.

Diverting superannuation money from retirement into Australians’ wallets today is not without cost. But it will put money into the hands of people when they most need it. And they need it now.

Footnotes

1. Affected workers will be able to apply online through myGov to access up to $10,000 of their superannuation before 1 July 2020. They will also be able to access up to a further $10,000 from 1 July 2020 for approximately three months.

2. At age 67 years. Assumes nominal returns of 6.65 per cent after fees; inflation of 2.5 per cent; and an effective tax on super earnings of 8 per cent in the accumulation phase. See our Balancing Act working paper for our full set of retirement modelling assumptions.

3. All figures adjusted to today’s dollars assuming inflation of 2.5 per cent. All other retirement income modelling assumptions as per Grattan Institute’s 2020 Balancing Act working paper.

4. In this scenario we assume that workers effectively withdraw $30,000 from superannuation, rather than $20,000. We also do not model the effect of the unemployment that most people who access their super will be experiencing. However, previous Grattan work shows that even 5, 10, or 15-year career breaks have only a small effect on retirement incomes. Lower super savings are offset by higher retirement incomes.

Adapated from Coates, Brendan and Nolan,  Jonathan 2020, COVID-19: Giving people early access to their super is the right move Grattan Institute

But after meeting Prime Minister Scott Morrison last night, the two states backtracked and for now will only close pubs, clubs, cinemas, nightclubs, and restaurants. Schools will be closed in Victoria and the ACT, and parents will be encouraged to keep their children home in NSW.

State governments should stick to their guns and move more quickly to shutting down more non-essential businesses and activities.

Our best hope to limit the long-term economic damage and to save the lives of our friends and families is to do everything we can to reverse the spread of this virus.

Coronavirus is growing exponentially in Australia – with a sufficiently broad shutdown, it should fall exponentially as well. Choosing Endgame C now means that the shut-down will be much shorter than if we wait another week.

Australian COVID-19 cases up to Monday March 23

The goal should be to all but remove coronavirus from the community as soon as possible.

Modelling shows that “flattening the curve” is unlikely to save the health care system, and it definitely won’t save the economy.

Business cannot return to normal while this disease festers. But once infections are very low, tracking and tracing them becomes feasible, particularly if we upgrade existing systems.

Whatever restrictions are implemented, the challenge for the states will be to set community expectations so there is all but universal compliance.

The most effective public health messages are clear and simple. When people are told that it is too dangerous to go to a café but they are fine to get a haircut, they are right to be confused.

If the messages are contradictory, many people will ignore them, and we will waste our best chance to contain this virus.

Legal enforcement also helps to send the message, and Victorian Premier Daniel Andrews should be commended for announcing that 500 police will be knocking on doors to check that people are following self-isolation rules.

We’ll need control of our borders

The federal government must also step up and take control of our borders. If they cannot track every new entrant to ensure compliance, the borders should be closed to passengers completely, or quarantine should be enforced in airport hotels.

Australians may be complacent about the spread of coronavirus because we so far have had fewer cases than the UK, the US or Italy. But Australia is a smaller country; we need far fewer cases to create a crisis.

We are only just behind the UK when it comes to coronavirus cases per person, and only a couple of days behind France, Germany and the US. We are in a similar position to Italy three weeks ago.

Our biggest advantage is that we are testing more people than these countries, and growth of infections is a little slower, but there are no signs yet that we are changing the trajectory of our infection rates.

But every country is different – what works in one climate or culture might not be as effective in Australia. The risks of doing too little too late are high. The risks of doing too much are relatively small.

If Endgame C succeeds, Australians society will be able to slowly return much more closely to normal functioning after eight to twelve weeks.

South Korea, Singapore, and Hubei province in China have successfully implemented Endgame C – and their infection rates have fallen.

Economic life is reappearing, and they now have the benefit of a public health workforce that can focus laser-like attention on any new outbreaks to prevent widespread community spread.

With Endgame C, Australians can have hope for a brighter future.

by John Daley, Jonathan Nolan

Published in The Conversation, March 24 2020

But after meeting Prime Minister Scott Morrison last night, the two states backtracked and for now will only close pubs, clubs, cinemas, nightclubs, and restaurants. Schools will be closed in Victoria and the ACT, and parents will be encouraged to keep their children home in NSW.

State governments should stick to their guns and move more quickly to shutting down more non-essential businesses and activities.

Our best hope to limit the long-term economic damage and to save the lives of our friends and families is to do everything we can to reverse the spread of this virus.

Coronavirus is growing exponentially in Australia – with a sufficiently broad shutdown, it should fall exponentially as well. Choosing Endgame C now means that the shut-down will be much shorter than if we wait another week.

Australian COVID-19 cases up to Monday March 23

The goal should be to all but remove coronavirus from the community as soon as possible.

Modelling shows that “flattening the curve” is unlikely to save the health care system, and it definitely won’t save the economy.

Business cannot return to normal while this disease festers. But once infections are very low, tracking and tracing them becomes feasible, particularly if we upgrade existing systems.

Whatever restrictions are implemented, the challenge for the states will be to set community expectations so there is all but universal compliance.

The most effective public health messages are clear and simple. When people are told that it is too dangerous to go to a café but they are fine to get a haircut, they are right to be confused.

If the messages are contradictory, many people will ignore them, and we will waste our best chance to contain this virus.

Legal enforcement also helps to send the message, and Victorian Premier Daniel Andrews should be commended for announcing that 500 police will be knocking on doors to check that people are following self-isolation rules.

We’ll need control of our borders

The federal government must also step up and take control of our borders. If they cannot track every new entrant to ensure compliance, the borders should be closed to passengers completely, or quarantine should be enforced in airport hotels.

Australians may be complacent about the spread of coronavirus because we so far have had fewer cases than the UK, the US or Italy. But Australia is a smaller country; we need far fewer cases to create a crisis.

We are only just behind the UK when it comes to coronavirus cases per person, and only a couple of days behind France, Germany and the US. We are in a similar position to Italy three weeks ago.

Our biggest advantage is that we are testing more people than these countries, and growth of infections is a little slower, but there are no signs yet that we are changing the trajectory of our infection rates.

But every country is different – what works in one climate or culture might not be as effective in Australia. The risks of doing too little too late are high. The risks of doing too much are relatively small.

If Endgame C succeeds, Australians society will be able to slowly return much more closely to normal functioning after eight to twelve weeks.

South Korea, Singapore, and Hubei province in China have successfully implemented Endgame C – and their infection rates have fallen.

Economic life is reappearing, and they now have the benefit of a public health workforce that can focus laser-like attention on any new outbreaks to prevent widespread community spread.

With Endgame C, Australians can have hope for a brighter future.

Adapted from

Daley, John and Nolan, Jonathan 2020 COVID-19: We’re running out of time to use Endgame C The Conversation.

In 2013, the OECD poverty measure found that 26 per cent of Australians aged 65 and older suffered income poverty in 2013, compared to 13 per cent across all OECD countries.

But there are a number of problems with the OECD measure. It does not take into account drawdowns on savings outside super, and does not adequately account for housing costs. And like other poverty benchmarks, small changes in income can produce radically different rates of poverty.

Old-age poverty in Australia under the OECD’s relative measure is extremely volatile from year to year, because the Australian Age Pension sits right on the edge of the OECD’s poverty benchmark. For example, old-age poverty in Australia apparently fell sharply from 22 per cent in 2011 to 12 per cent in 2018.113 But the big apparent shift merely reflected the maximum rate of the Age Pension (including related supplements) oscillating around the benchmark of 50 per cent of median incomes. The minimum pension in many other countries is much lower and in some is available only to people who have been employed for most of the time while they were of working age. As a result, Australia has far fewer retirees in severe poverty whose income is much less than the OECD’s benchmarks.

This clustering close to the benchmark also means that outcomes on the measure depend a lot on somewhat arbitrary definitions. For example, the apparent poverty rate in 2017-18 changes from 14 per cent to 25 per cent, depending on how households with different family sizes are compared. Even then, the income poverty measure can be misleading. The 14 per cent of senior Australians classified as living in poverty on the ABS preferred definition are often people of significant means who are ineligible to receive a maximum-rate Age Pension and whose drawdowns of existing savings are not counted as income. More than half of over-65s classified as living in poverty in 2017-18 based on the ABS preferred definition were among the wealthiest half of all retirees.

In many other OECD countries the depth of poverty for older-aged people is much greater - meaning that instead of straddling the poverty line, low-income people sink far below it.

A more reliable measure of poverty for retirees, especially renters, is income poverty after housing. This is the proportion of households that have less than half the median disposable income once housing is taken into account. To account for housing, this measure uses a standard ABS estimate of the value home-owners get from not having to pay rent. The measure raises the income of home-owners significantly, while exposing the tough economic circumstances faced by many renters. On this measure, about 10 per cent of Australians aged 65 and older live in poverty after taking account of their housing costs, but results vary wildly by housing tenure. About 60 per cent of retired Australians who rent and live alone are in poverty after housing costs. In contrast to less than 10 per cent of retired homeowners without a mortgage.

Retirees in Australia also feel financially comfortable. Most retirees have paid off their mortgage, and no longer have the financial stress of bringing up children. They are far less likely to suffer financial stress such as not being able to pay a bill on time, and are more likely than working-age households to say they feel financially comfortable Just as retirees are less stressed about essentials, their discretionary expenditure is also less financially constrained than that of working-age people. Retirees are less likely to miss out, due to cost, on things like taking a holiday. Retirement is a particular relief for low-income earners, whose income typically increases in retirement with access to the Age Pension.

Rather than running out of money each week and dipping excessively into savings, higher-income Australians maintain their nest egg well into retirement. Most retirees never spend a large part of the savings that they have on the day they retire. Many retirees seem reluctant to draw down on their capital, and instead live on the income their savings generate. While precautionary savings may explain part of retirees’ savings behaviour (Section 1.6 on page 17), the combination of a lack of spending and high rates of financial satisfaction suggests that retirees’ needs are being met.

Home-owning retirees receive a significant financial windfall because most of the value of their home is effectively excluded from the Age Pension assets testBut . For pensioners who don’t own their own home, Rent Assistance is a fortnightly payment that can help bridge the gap.

But rent assistance is not high enough. The maximum Rent Assistance payment is indexed in line with CPI, but rents have been growing faster than CPI over the long term. Between June 2003 and June 2019, CPI increased by about 46 per cent, while average (quality adjusted) rents increased by about 65 per cent.

The Age Pension is sufficient for most retirees to have an income above most measures of poverty. But it is not sufficient for renter households. Commonwealth Rent Assistance should be increased by about 40 per cent, to overcome its decline relative to average rents over the past two decades. This would be the most targeted way to alleviate poverty among retirees.

This is an adapted exerpt from Coates, B. and Nolan, J. (2020). Balancing act: managing the trade-offs in retirement incomes policy. Grattan Institute.

Our research shows the super increases aren’t necessary. For most Australians, retirement incomes are already adequate. Since higher super contributions will come at the expense of wages, the scheduled increases should be abandoned.

But there are big problems the review will need to confront.

Here are five changes that would tackle them.

1. Boost rent assistance

While most Australians are comfortable in retirement, the system is failing too many poorer Australians, especially low-income women and retirees who rent.

Senior Australians who rent privately are more likely to suffer financial stress than homeowners or renters in public housing. And it will get worse because young Australians on lower incomes are less likely to own homes than in the past.

The government’s priority should be boosting rent assistance, which has not kept pace with rent increases. Raising rent assistance by 40%, or roughly A$1,400 a year for singles, would cost just $300 million a year if it applied to pensioners, and another $1 billion a year if extended to other renters.

A common concern is that boosting rent assistance would lead to higher rents. But that’s unlikely: households would not be required to spend any of the extra income on rent, and most would not.

2. Ease the age pension asset test

While retirement incomes are adequate for most retirees, the age pension assets test excessively penalises people who save more for their retirement.

Before January 1, 2017 retirees with assets above the threshold lost $1.50 of pension per fortnight for every $1,000 of assets above the threshold. In 2017 the Coalition lifted the threshold but also lifted the withdrawal rate to $3 of pension per fortnight for each $1,000 of assets.

The changes resulted in very high effective marginal tax rates on retirement savings, so much so that a typical worker who saves an extra $1000 at age 40 increases their retirement income by only $25 each year, or $658 over 26 years of retirement, which is a negative return on money saved for decades.

The age pension withdrawal rate should be cut to $2.25 per fortnight for each $1,000 of assets above the threshold. This would cost the budget about $750 million a year.

For middle and high-income workers, this change would have a bigger impact on retirement incomes per government dollar expended than boosting compulsory super.

3. Boost Newstart

Newstart, together with the disability support pension, provides an important safety net for Australians who are unable to work right through to retirement age.

Yet while the age pension and disability support pension are indexed to wages, Newstart is not. It only climbs in line with inflation. It should be increased by $75 a week and then indexed to wages going forward.

This would cost a lot but it would help the growing legions of older Australians, many of them women, who find themselves among the long-term unemployed in the years leading up to retirement, or are forced to retire early. And it would lift many more younger Australians out of poverty.

4. Include the home in the pension assets test

Falling rates of home ownership mean we are at risk of creating an underclass of retirees who rent.

And our retirement incomes system makes this worse by favouring homeowners over renters. Once a person is retired, their home is treated differently to their other assets. Which is why $6 billion in pension payments go to people with homes worth more than $1 million.

It’s time for more of the value of the family home to be included in the pension assets test. Counting more of the home above some threshold (such as $500,000) would be fairer and would save the budget up to $2 billion a year.

No pensioner would be forced to leave their home. Pensioners with valuable homes could continue to stay at home and receive the pension under the Government’s pension loans scheme, which recovers debts only when homes are eventually sold.

5. Fix super tax breaks

Superannuation tax breaks cost a lot – tens of billions each year in foregone revenue, with half the benefits flowing to the top one fifth of income earners, who already have enough resources to fund their retirements.

And the costs are set to climb further as super balances climb. The cost of the earnings concessions alone is set to climb from $17.4 billion to $20.8 billion over the next four years.

Three reforms would keep them in check.

• Voluntary contributions from pretax income should be limited to $11,000 a year. This would save the budget about $1.7 billion a year.

• Contributions from post-tax income should be limited to $250,000 over a lifetime, or to $50,000 a year. It won’t save the budget much in the short term, but in the longer term it will plug a large hole in the tax system.

• Earnings in retirement – currently untaxed for people with superannuation balances less than $1.6 million – should be taxed at 15%, the same as super earnings before retirement. Doing so would save the budget about $2 billion per year at first, and much more in future.

These changes to super taxes free up money to help Australians who need help without hurting the retirement prospects of middle Australians.

Australia’s retirement incomes system works well, but there are things that need fixing.

The reforms we propose would make retirement fairer, save taxpayers’ money, and ensure that all Australians can enjoy a comfortable retirement free from poverty.

by Brendan Coates, Jonathan Nolan

Published at The Conversation, March 2 2020

People who are forced to retire early, such as carers or those with a disability, used to be able to rely on the Carer Payment or the Disability Support Pension. But as shown in in the graph below, recent changes to eligibility requirements for the Disability Support Pension mean fewer older Australians access the scheme then previously. And Newstart remains inadequate, having not been increased for the past two decade

Forcing all Australians to self-insure by saving even more for retirement, just in case they retire early, makes them poorer while working and is simply a recipe for larger inheritances.

Eligibility requirements for the Disability Support Pension were tightened in 2012, to the detriment of many near-retirees with musculoskeletal health problems. In 2009, about 12 per cent of 55-64 year-olds were on the Disability Support Pension. By 2017 that number had fallen to 9 per cent. The decline coincides with an increase in the number of older people on Newstart.

Newstart has also become woefully inadequate as a safety net for unemployed Australians. Unlike wages or pensions, Newstart has not increased in real terms in more than 20 years. While the Age Pension is indexed to wages, Newstart only increases with inflation. This has ‘squeezed’ the living standards of people living on Newstart relative to the rest of the population.59 A typical single person on Newstart receives just $39 a day, about 18 per cent of average (male) earnings.

Households of working age receiving welfare payments – primarily Newstart – are under much more financial stress than households receiving other welfare payments.

This is an adapatation/exert based on research in

Coates, B. and Nolan, J. (2020). Balancing act:managing the trade-offs in retirement incomes policy. Grattan Institute

Australians tend to spend less after they retire. Even the wealthy eat out less, drink less alcohol, and replace clothing and furniture less often. Spending tends to slow at around the age of 70, and decreases rapidly after 80.

This fall in overall spending is mainly a result of lower spending on transport, recreation, food and furnishings. Retirees who own their home tend to have paid off the mortgage, and retirees no longer need to spend money on children or on work-related expenses.

Pensioners also spend less because they get discounts on council rates, car registration, electricity and gas bills, public transport fares, and pharmaceuticals. Public transport concessions apply to all retirees – not just pensioners. Retirees’ spending also tends to be lower because they have more time, and so cook at home more and eat out less.

They spend more on health care as they age, but Medicare largely shields them from the full costs. The modestly higher out-of-pocket costs they do pay are mainly due to rising premiums for private health insurance.

Not only do most retirees not draw down their savings throughout retirement, many add to them. One recent study found that the typical pensioner still had 90 per cent of what he or she retired on after eight years. Retirees say they actually feel more comfortable financially than the Australians younger than them who are still working. And retirees are less likely to suffer financial stress such as not being able to pay a bill on time. Even retirees who rent are less financially stressed than people who are working and renting.

International studies make similar findings. Reports using the British Family Expenditure Survey and The American Income Dynamics and the Consumer Expenditure Survey both found spending decreases into retirement. Another prominent US study found that real spending falls by around 1 per cent each year in retirement.

What this means for retirement incomes policy

Our analysis using the HES shows that retirees’ spending needs fall as they age, but not because they can’t afford to spend more, or in a way that leaves them unsatisfied. The fall in spending appears to reflect declining recreation and other discretionary spending as retirees age and their health declines. Meanwhile governments foot most of the bill for retirees’ rising health and aged care costs.

This means that calculations about the adequacy of retirement savings ought to be based on whether they are enough to maintain buying power (at best). Yet much of the existing Australian research assumes that retirees need to save enough to enable their incomes to keep climbing throughout their retirement, in line with general wage growth. Implicitly, these studies assume that a retiree needs to spend 22 per cent more at age 90 than at age 70, after accounting for inflation.

The result is that we overestaimte how much most people need to save to have a happy retirement.

Australia’s retirement income system doesn’t work for everyone. Senior Australians who rent in the private market are more likely to suffer financial stress than homeowners, or renters in public housing. Falling rates of home ownership among younger Australians mean more Australians will rent in retirement. A growing number of older Australians, including many women, are at risk of poverty and homelessness in retirement. And Australians who suffer from major shocks, such as forced early retirement due to illness, are not as well protected by our income support system as they should be.

These are the real challenges to ensuring our retirement incomes system lives up to its promises to all Australians. But more superannuation is not the answer.

Excerpt from Coates B, Nolan J. Balancing Act: Managing the Trade-Offs in Retirement Incomes Policy (Submission to the Retirement Income Review). Grattan Institute; 2020.

In contrast to stamp duties, land taxes do not distort decisions about land use, provided they apply in a way that the landowner can’t avoid. For example, a constant-rate land tax applied to the unimproved value of all land prevents landowners from reducing their liability to such a tax by changing the way they use their land. The tax on a vacant block of land would not increase even after the block was developed.

And while the political economy of land taxes is challenging - the right design can help alleviate political barriers. The best way to reduce the political salience of land-tax is to avoid sending landowners a yearly bill. That way asset-rich but income-poor households can stay in their homes, by allowing them to defer paying the levy until they sell their property. Deferral arrangements are already available for seniors paying council rates in South Australia, Western Australia, and the ACT.

A recent paper by Prosper Australia argues that deferred land tax should be available to all households, but interest on any deferment should be charged at market rates. An independent statutory body could then hold any deferred land tax liabilities, and provide revenue to government in the form of a dividend to the government every year.

The section above was adapated from Coates B, Nolan J. NSW Should Swap Stamp Duties for a Broad-Based Property Tax. Grattan Institute; 2019.

There’s a million ways a land tax could be phased in, but one I particularly like is by making it revenue positive from the beginning. Start by calculating how much land tax every homeowner would have paid had there always been a land tax. If the stamp duty system was giving them an effective ‘discount’ on their land tax bill, they should have to pay the difference on any land tax that accrues from the start date of the law. The figure below shows how it would work:

The extra revenue generated could be used to placate voters and ensure a just transition. Over time as revenues increased governments could make the decision to lower stamp duties or even lower other taxes.

Land tax ‘reforms’ are fairly criticised as a way for conservative governments to starve the beast and reducing overall tax revenue. But with this reform we could have the best of both worlds - increasing the most efficient tax while doing whatever is politically expedidient with the rest of the tax base.

While there are complications in cross-country comparisons of superannuation fees, a more complete picture shows Australian superannuation fees are not among the lowest in the world. And more direct measures of fees suggest they remain high by international standards.

The OECD’s figure misses much of the story

The latest OECD Pension Markets in Focus report puts superannuation fees in Australia at around 0.5 per cent of assets under management.

But in fact very there are relatively few superannuation funds where the typical member is paying fees of 0.5 per cent or less of their super balance.

So where does the OECD figure come from?

The OECD’s estimate of Australian superannuation fees come from APRA’s annual Superannuation Bulletin report, which asks super funds the total amount they charged in fees to members each year. The trouble is that the APRA data used by the OECD misses much of the story.

In total just $9.4 billion in direct fees were paid to APRA-regulated funds in the year to June 2018, or around 0.5 per cent of total assets.¹

Yet the Productivity Commission (PC) also found (p.163) that many superannuation funds do not report the fees they pay to companies who help to manage members’ investments. And the PC took APRA to task for the fact that at least a quarter of all super funds, and especially higher-fee retail funds, didn’t report all their fees. APRA has begun to take steps to improve its data collection practices, but any changes will take time to come into effect.

In total the PC estimated that APRA data for 2017 missed $7 billion in indirect investment costs and underreporting of direct fees among APRA-regulated funds with MySuper products.²

The OECD data also excludes self-managed superannuation funds, which tend to have higher costs than APRA-regulated funds.³

Australia’s superannuation fees are much higher than the OECD’s figures

In contrast, ASIC requires that superannuation funds disclose all their fees to new members in product disclosure statements.⁴

When the PC collated all of the fees advertised by superannuation funds in 2017, it found average fees were 1.1 per cent – more than two times higher than the initial OECD’s estimate for fees as a share of assets. Further, the industry’s own estimates typically put total superannuation costs at or above 1 per cent of total assets under management. For example, Rainmaker estimate total superannuation fees of $31.8 billion in 2018-19, or 1.1 per cent of total superannuation account balances.⁵

The OECD acknowledges these issues in its report, stating:

“Data may underestimate the actual charges on the pension pot (sic) paid by members in some countries through indirect costs reducing investment returns for instance.”

Superannuation fees of 1 per cent might seem small, but the bill must be paid on the entire balance of your superannuation account every single year. Small changes in fees can really add up.

A young person today paying fees of 1 per cent instead of 0.5 per cent, will lose around 10 per cent of their retirement nest-egg. ⁶

High fees would not matter so much if we were confident that high-fee super funds were earning their keep. But Grattan’s 2015 report Super Savings found that funds that charge high fees were also some of the worst performers, delivering poor value for money for their members.

The PC similarly concluded (p.15):

“While some may be receiving exceptional investment returns or member services, the evidence indicates that funds that charge higher fees tend to deliver lower returns, once both investment and administration fees have been netted off. High fees also persist over time.”

More direct comparisons also suggest Australia has relatively high superannuation costs

Of course international comparisons of pension system costs are complex. As others have noted, differences in regulatory structure, taxation, and reporting requirements will all affect the results. Meanwhile underreporting of fees and indirect costs may occur in other jurisdictions too.

To side-step these factors, the PC also directly compared the investment management costs of Australian super funds against international peers by asset class. It found that for most asset categories, Australians are paying much higher fees than pension funds in the US or Europe.⁷

In fact when it came to Australia’s superannuation fees the Productivity Commission concluded (p.158):

“Superannuation fees in Australia are higher than those observed in other OECD countries. This may be partly because Australian funds face higher expenses. While international comparisons are not straightforward, there is evidence (by asset class) that Australian investment management costs are generally high by international standards, including for significant asset classes (such as equities and international fixed income).”

On any reasonable measure, Australian superannuation costs still appear high by international standards, and are certainly far from being the lowest as the latest OECD data suggests.

Footnotes

1. APRA Superannuation Bulletin 2018, Table 6.

2. See Figure 3.19 of the Productivity Commission’s final report on Superannuation: Assessing Efficiency and Competitiveness. 

3. See Figure 3.22 of the Productivity Commission’s final report on Superannuation: Assessing Efficiency and Competitiveness. 

4. Although as the PC noted (p.155), they may not capture their full extent for all funds and products.

5. Rainmaker also estimated total fees of 1.17 per cent of account balances across APRA-regulated super funds alone. Rice Warner estimated total superannuation fees of $22 billion in the year to June 2017, or 1 per cent of assets. For a comparison of the various estimates of total fees as a share of superannuation assets, see Figure 3.3 of the Productivity Commission’s final report on Superannuation: Assessing Efficiency and Competitiveness.

6. Assuming a 40-year contribution period, real wage growth of 1.8 per cent p.a., and real portfolio returns of 5 per cent. See Figure 4 of Grattan Institute’s 2014 report, Super Sting, for further details.

7. The PC found Australian funds incur much higher investment costs for domestic and international equities, and international fixed interest, compared with most other countries in the dataset. Australian funds also pay marginally higher costs for cash and domestic fixed income on average. By contrast, the investment management costs for Australian superannuation funds are relatively comparable to international pension funds for unlisted assets (including unlisted infrastructure, unlisted property and private equity). See Productivity Commission final report Superannuation: Assessing Efficiency and Competitiveness, Figure 3.2 for further details.

Adapted from: Nolan J, Coates B. The curious case of the missing superannuation fees. Grattan Blog. Published 2019.

The right way to look at teachers’ pay is to ask whether it is high enough to attract the people we want and need into teaching. In Australia the answer is no.

Australia is a rich country. Our incomes are high in comparison with many other countries. But our education system is competing with other industries to attract the best talent.

A more meaningful international comparison is to look at how teachers’ pay lines up with the incomes of other professionals in their own country. While Australia’s pay for young secondary teachers relative to other professionals stacks up well internationally, our pay for secondary teachers in their mid-thirties and their forties is below the international average.

Chart 1. Teachers quickly fall behind their peers in other professions
_{Total yearly personal income of full-time workers holding a bachelor degree, by field of study, 2016, at the eightieth percentile of the income distribution}

_{Notes: Includes people who studied a teaching degree but now work as principals or outside teaching. Four-digit fields of education chosen because they have the highest median ATAR. “No degree” includes all levels of education below bachelor.
Sources: Australian Bureau of Statistics; Grattan analysis}

In other words, Australia’s young teachers are paid well, but the pay scale is flat. Teachers’ pay at the top is only 1.4 times as much as their starting salary, below the OECD average. An Australian classroom teacher’s pay stops rising within ten years, whereas the incomes of his or her peers in other professions keep rising well into their thirties and forties.

Chart 2. Since the 1980s, teachers’ pay has fallen well below pay in other professions
_{Average teacher salaries as a percentage of all professionals, 1986 to 2018}

_{Notes: Salaries measured as the average weekly cash earnings of full-time non-managerial adult employees. In 2006 the ABS started including salary-sacrifice income and changed the definition of “professional.” For years 2010–18 aggregate incomes for all professionals were not available, and so they were calculated using the weighted average of all professional occupations. The weights were derived from the number of full-time workers in each occupation and gender category in the 2016 Census.
Sources: Australian Bureau of Statistics; Andrew Leigh, “Teacher Pay and Teacher Aptitude.”}

But average pay is not the only factor. More important — especially for attracting young high achievers — is the potential to earn a very high income if you do a good job. Bright young Australians have a good chance of earning a high income in many professions, but teaching is not one of them. More than a third of engineering or commerce graduates in their forties working full-time earn more than $3000 a week ($156,000 a year). But for graduates with a teaching degree, that share is only 2.3 per cent.

A Grattan Institute survey of almost 1000 young high achievers (aged eighteen to twenty-five and with an ATAR of 80 or higher) shows that 70 per cent are open to becoming teachers. But university enrolment data show that, when it comes to the crunch, only 3 per cent of high achievers actually choose teaching for their undergraduate studies. Our survey shows that a major barrier is teachers’ pay.

Australia’s young bright people are being asked to make a big financial sacrifice if they choose teaching as their career. They know they will be staring down the barrel of a lifetime of pay much lower than that of their classmates who choose degrees such as engineering, science and law.

Chart 3. High achievers say teaching falls short on intellectual challenge, and pay
_{Young people who state that a career in teaching is more likely to provide a given attribute than their chosen occupation}

_{Note: Career attributes are ordered top-to-bottom from most to least important. The data in the chart show the difference (teaching minus chosen occupation) in the percentage of respondents who answered that a given career was likely or very likely to provide each attribute.
Source: Grattan Institute survey of high-achieving young Australians.}

This is a big problem. Teacher “smarts” matter. Evidence shows that people with strong academic records do a better job teaching in the classroom. Yet over the past forty years, teachers’ pay has been declining in Australia relative to other professions, and with it fewer bright young people have chosen teaching as a career.

The fact that only 3 per cent of high achievers choose to study teaching compares very unfavourably with 19 per cent for science, 14 per cent for health, and 9 per cent for engineering. Over the past decade, high achievers’ demand for teaching fell by a third — more than for any other undergraduate field of study.

Australia’s school education system is not in a good place. Our kids are falling behind. In international tests, the typical Year 9 student performs at a much lower level than a similar-age student in 2003 — lower by twelve months in maths and nine months in reading.

We can never know for sure what is causing our students to fall back, but it’s difficult to ignore the fact that Australia’s test-score decline has coincided with the retirement of many of the teachers who were recruited when salaries were much more competitive with other professions.

If Australia is to catch up, bold action must be taken. A new Grattan Institute report, Attracting High Achievers to Teaching, proposes a $1.6 billion reform package for government schools to double the number of high achievers who choose to become teachers. It would increase the average ATAR of teaching graduates from 74 to 85 within the next decade.

The package includes $10,000-a-year scholarships for high achievers who take up teaching, and new career paths for leaders of the profession, with pay of up to $180,000 — about $80,000 more than the current highest standard pay rate for teachers.

If governments were to implement this blueprint, it would send a strong message to Australia’s best and brightest: if you want a challenging career where expertise is paid accordingly, choose teaching. In the long term it would pay for itself many times over, because a better-educated population would mean a more productive and prosperous Australia.

by Julie Sonnemann, Jonathan Nolan

Published by Inside Story, Thursday 25 September 2019