Sunday, June 07, 2020

Five ways Melbourne's public transport is anti-poor

Back in March I asked whether the current public transport service offering disadvantaged women relative to men. You can read that item here

An important supplementary question was whether any discrimination was because women were women or because they, on average, received lower incomes than men. The answer is that it depends. 

Take near-ceiling grab bars on trains, for example. That these are unreachable for more women than men indicates an aspect of carriage design whose effect is more sexist than classist. This is because sex is a bigger predictor of average height within a population than class. Variations by national origin appear greater than class but even here there is a wealth dimension as shown when comparing North Koreans against the Japanese and the rapidly growing South Koreans

Fixing this just by lowering the bars isn't great; tall people (mostly men) might bump heads more. Straps could help though apparently they were removed about 20 or 30 years ago as vandals were holding onto them to kick out carriage windows. More vertical poles is another possibility though placement should enable efficient access and egress to reduce dwell times at stations. 

So much for this example of (likely unintentional) sexism in one small aspect of public transport design. What about discrimination based on class? That is the factors like income, education,  employment, social status, etc that differentiate people from one another.   

Is public transport an enabler of opportunity in our cities? Or does it hold people back? Do decisions on infrastructure, routes, timetables and fares systematically discriminate against those with less money, less formal education, less social capital and in inferior jobs? If so, how? And are things being done about redressing that in an organised way like we see for disabled access (DDA compliance/accessibility programs) and women (Women in Transport program)? 

Classism - the unspoken ism

I'll get back to those questions later. First of all I should mention that there doesn't seem much talk about public transport policy and classism in Australia. Or maybe there is, but unlike the televised board meetings we see from some North American transit authorities, our Department of Transport keeps to itself. They may well think about the environmental and distributive justice effects of their priorities but there's no public evidence of this. The department's last annual report indicates an overwhelming preference for infrastructure over service and its planning if mentions are anything to go by. The best principles and intentions won't help if implementation is glacial.  

Debate is more open and developed in US cities. Compared to ours their class divisions are greater, neighbourhoods are more racially segregated and public transport worse. While underserving transit's best customers with crowded city buses, American cities sometimes build expensive low frequency commuter rail or short distance light rail (often with funding from federal grants or referendum votes). These often have a high subsidy per passenger with benefits skewed towards wealthy areas or developers. 

Americans' municipal management of public transport doesn't help given sprawl across historic boundaries and job suburbanisation. Also, because the funding base of transit services in the US often depends on volatile sales taxes, recessions such as 2008's GFC can lead to large service cuts of the type we haven't seen in Australia since the early 1990s. 

For some background on US transport politics read up on LA's Bus Riders Union and similar groups. Spending on bus services in Hispanic and black areas rather than commuter rail from mainly white areas, improved access to jobs, fare cuts and hiring bus drivers rather than police are some of their aims. They want civil rights and distributive justice rather than merely procedural justice. In the US civil rights and urban transit have been intertwined at least since Rosa Parks.

Bus service distribution, patronage and demographics in Melbourne

Our social divisions are less stark than in US cities. But anyone who rides the buses throughout Melbourne will spot echoes of US themes. For example relatively sparsely populated areas with good service but low per hour bus usage (eg Brighton, Greensborough, Eltham, Warrandyte, Yarrambat, Emerald) are demographically whiter than average. Buses may typically record 10 to 20 passenger boardings per bus service hour on weekdays. They might overlap other routes or run until midnight at good frequency despite few users. Some examples presented here and here

In contrast, lower income suburbs with underserviced and/or heavily used services often have more east Asian, Indian, middle eastern or African faces. Buses are busier, with 25 to 60 weekday boardings per bus hour common. Some routes may be daytime only and not run Sundays. A cost-neutral 'colour-blind' network review would add trips to these productive routes at the expense of the quieter services mentioned before. 

The chart below gives a snapshot of services in some low and high non-English speaking middle, outer and fringe areas. Only some suburbs are listed but the relationship could hardly be clearer.  

Polite people would never accuse the 'nice' degreed middle-class inner-suburban political adviser and bureaucrat types around the Department of Transport as being racist. However the consequence of the neglect of bus resourcing and network reform has distributional injustices that no policy or planning professional should run away from. 

The specifics 

Let's get back to class. A good starting point is these maps prepared by Nathan Lambert. They show the twenty areas of largest disadvantage in the state. In descending order those in Melbourne are Braybrook-St Albans, Springvale-Dandenong, Broadmeadows-Thomastown, Heidelberg West- Reservoir, central Melton, Frankston, northern Werribee, Rosebud and Hastings. There are also some inner suburban public housing towers. The largest clusters are associated with majorities of non-English speakers but this is not so for smaller or more distant clusters like Melton and Rosebud. Also see this Age report and map for another representation.

It's worth being aware of the large number of 'working poor' that just miss Lambert's classifications. They might live near but not in the twenty clusters. Because they have at least some work they aren't the poorest of the poor. But from a transport point of view this group likely contains many one car families whose ability to practically get and hold extra jobs is influenced by available transport choices. Large concentrations of this diverse group, which I wrote about here, are found in areas like Tarneit and Craigieburn, and, to a lesser extent, parts of the south-east (with higher car ownership).

Here are five ways where public transport may be seen to work disproportionately against the poor. This can reduce opportunities since when public transport is limited it is these people who have least access to alternatives that are dignified and safe. For example testing friendships by asking family and friends for lifts too often, hitch-hiking or walking along roads with dangerous large roundabouts at uncrossable intersections.

1. Good infrastructure and service levels distant from where low income people live.

Which areas have the most train and tram infrastructure? It's generally the inner suburbs before the lines either finish or fan out to be distant from one another. Widespread inner suburban gentrification and outer area degentrification has made it more likely that higher income people have better access to frequent trams and trains than those further out.

Hence there's an infrastructure bias against poor people. You might argue this is inevitable. Infrastructure such as railways improves accessibility which creates demand and inflates land values. High prices and rents then makes the area unaffordable to poorer people who move to places more distant or less accessible. The solution is to extend infrastructure in poorer areas and use what we have more intensively.

This brings us to service levels. Up to a certain limit they're quicker and easier than infrastructure to expand. Provided they boost driver training and start timetable development soon after gaining office, a new government can introduce at least five rounds of train timetable upgrades in their first four year term (each separated by six months). This presents a great opportunity for a service-oriented minister to make a major difference, even in years when capital works budgets are lean.

That's the theory. In practice, due to political inertia, timetables have been almost as permanent as built infrastructure. Hence they can perpetuate old inequalities almost as much.

As an example, the south and east contain the largest clusters of wealthier suburbs. Their suburban lines all get a 10 to 15 minute interpeak service to Sandringham, Frankston, Dandenong, Glen Waverley, Alamein and Ringwood.

In contrast the north and west are all every 20 minutes including to low income pockets such as Thomastown, Broadmeadows, St Albans and Werribee. This is despite demographics such as low labour force participation being conducive to high off-peak patronage that justifies a better timetable especially during these virus-separated times.

While the south and south-east got interpeak train frequency increases in the 1990s and 2000s, the north and west remain stuck with 1970s-style 20 minute interpeak frequencies. The reason comes down to Liberal ministers such as Brown and Mulder skewing service increases to southern and eastern seats they hold and Labor ministers such as Allan and Horne, spoiled by high margins, taking western and northern seats for granted and not spreading eastern suburbs-type upgrades to 'their' areas.

What about buses? The big SmartBus roll-out about 10 years ago was a giant step forward. Disadvantaged areas around Frankston, Dandenong and Springvale gained immensely. However it also perpetuated existing east/west inequalities. Six of the nine SmartBus routes start in and remain in the eastern suburbs. Three penetrate the northern suburbs. Of these just one makes it to the west (Altona, via a non-optimal, duplicative corridor). None serve high-demand/high-growth areas such as Tarneit and Craigieburn, although two serve slow-growth areas around Eltham/Greensborough (often duplicating other routes).

Despite the above good things have since happened. Outer parts of the north and west have got new or extended lines, for example to Wyndham Vale, Watergardens, Sunbury, South Morang and Mernda. Bus upgrades in areas like Wyndham, Brimbank, Epping North, Mernda, Cranbourne and Endeavour Hill have improved services from very limited to basic or better.

Also the benefits of the 2006 - 2010 minimum service bus upgrade program cannot be overstated. However it was never completed. Dozens of routes remain with short operating hours and 6 day service. Many are in poorer suburbs such as Springvale, Noble Park, Dandenong, Campbellfield, Thomastown and Rosebud. Limited operating hours and low frequencies pose a particular problem for poorer people since they are more likely to be at or be seeking jobs that require early, late or weekend work, sometimes at different locations.

To summarise the distribution of both infrastructure and service discriminates severely against the poor. The latter is cheaply and easily fixed but political interest in this has been patchy.

2. Routes and timetables don't suit jobs that low income people are more likely to work at.

The previous point dealt mainly where people live. They need good service from home to connect with people, services and opportunities outside it. This point is focused on the most important of these opportunities, namely jobs. Access to these is vital for social mobility, escaping poverty traps and reducing welfare dependence.

A city-based nine to fiver has strong radial train and tram lines in most directions. Where they don't have one near home there is likely a bus that will get them to one. Peak frequencies are mostly OK on train and tram lines so waiting shouldn't be too long (at least heading in to work) even if times don't perfectly match up. A lot of the jobs they are in are non-customer facing with flexible hours so it's no biggie if a cancelled train or bus makes you half an hour or more late.

Many things are different for the types of temporary, casual, part-time and sometimes low-skilled jobs that those who have been out of the labour market for some time might expect to get. For example they might be at a location without or distant from public transport. That's a particular issue with industrial area jobs where areas like Laverton North have no direct buses from major population centres such as Tarneit. Secondly working times might include early starts, evenings or weekends when few buses run. Thirdly flexibility might be on the employer's terms so that bus times don't match up with start and finish times. Unreliable service may even cost some their jobs or result in steep childcare late penalty fees. Fourthly interchanges will almost certainly be required and waits may be long if service is infrequent. All these affect many suburban jobs, especially those which involve manual labour or customer service.

Daytime CBD workers (who are mostly on good incomes) get close to the best the network offers while those taking buses to lower wage suburban jobs get close to the worst it offers in terms of coverage, operating hours and frequency. This reduces the financial benefits of taking a job as an extra car will almost certainly be required to travel to it even though it may not be needed for other purposes. Car ownership is lower in emerging dense outer suburbs such as Tarneit and Craigieburn than you might expect so this is a major issue for many families. There is currently no service-based network reform agenda to make public transport more useful for trips to suburban jobs. I discuss ways to economically build a Job Ready Network in a parliamentary inquiry submission here.

3. Fare policy and topping up.

You could argue the toss two ways on this. Low income people from Zone 2 commuting into Zone 1 get a pretty good deal. Local travel within Zone 2 (which is where most low income people live) attracts cheaper fares. And some working poor qualify for concession fares due to the low income Health Care Card.

However other aspects of the fare system, like lower per day rates for commuters on myki pass over casual travellers on myki money as well as the absence of off-peak fares make the system work against people on lower incomes. Myki pass discounts are great for full-time workers with predictable plans who can confidently pay a week's, month's or even a year's fare in advance. Less well off are more likely to have a variable work roster or lack the cash for large up-front fare payments. The sample size is small but it seems that bus passengers using myki money don't top up in large amounts. This tallies with work the Grattan Institute did that found that 10% of working households had less than $90 cash in the bank (It may be higher for all households). 

A more egalitarian fare structure would reduce discounts for longer term periodical tickets (especially for exurban commuters who these tickets grossly subsidise) with the extra money (since we know that peak travel is relatively fare inelastic) going to make off-peak myki money travel cheaper. The CBD Free Tram Zone should also go. Again these are cases where good economics and social equity point to the same policy reforms.  

It's part of their job for DoT/PTV people to decide what is important in published information. A not necessarily conscious anti-poor/anti Zone 2 mindset can cause bad judgments. For example fares for travel within Zone 2 (which form the staple of many low-income earners' travel) does not feature on information at each station.

This omission forces passengers to top up with an unspecified amount and trust 'the system' to deduct the right fare. Those making longer distance 1 + 2 trips are not treated in such a cavalier fashion with fares spelled out. There would be times at the end of the week when knowing how much to top up is important information when people are down to their last few dollars. Such life experiences would be either alien to or long-distant memories for the degreed well-paid Zone 1-based bureaucrats who makes these sorts of silly decisions.

Here's another thing. Full fare payers can buy a suitable myki at an unstaffed station (ie most of them) from a ticket vending machine. But concession holders can't. So if a concession passenger left their myki at home and wanted to do the right thing they'd have to pay double for a full fare myki. This policy is anti-senior, anti-child and anti-concession. It also increases the temptation to fare evade. PTV realised the issue and promised to fix this in 2015 when new-look mykis came out.

Over five years later you still can't buy a concession myki from a station ticket vending machine. What happened to Alan Fedda, the guy who made the promise? He did well. He became the Executive Director Franchise Operator Management at PTV and was also its deputy CEO. He was in PTV when Transdev Melbourne had its fleet management crisis which forced over 100 buses off the road on safety grounds. Was the Baillieu government to blame for awarding the business to the lowest bidder or should PTV have been more vigilant in its contract management? In any event it sounds like Fedda treated Transdev right because 18 months after the bus crisis he popped up as the boss of Transdev ferries in Sydney.

4. The system's hostility to luggage.

Steps and fare barriers can work against those carrying much more than a shoulder bag and laptop. Lower income people, especially if they do not own cars, are more likely to use public transport for moving large items including food shopping, clothing or even small appliances. Homeless often use shopping trolleys. Similar issues apply as for DDA access when it comes to boarding trains, trams and buses. Richer people generally don't find themselves in this situation since they are more likely to have access to a car or shell out for a taxi. The very existence of barriers is the result of a decision that although it's conventional here is not necessarily the case elsewhere, for instance some German cities with more open systems.

5. Passenger information.

I mentioned out-of-touch technocrats before. On-network maps and signage is something else they don't like. They have been responsible for omitting useful printed information from stations and other interchange points. A few years ago they removed maps from the back of stop-based bus timetables. Even new stations, like Frankston, get built with a paucity of multimodal network information.

A whole heap of classist assumptions go into these sort of decisions made with zero accountability or public scrutiny. For instance there is an assumption that public transport users are regular commuters making familiar trips near home or on a single known trip to work. Or that everyone carries a mobile phone with limitless data. These sorts of assumptions work against people who use public transport for a wide variety of occasional trips (such as people going to job interviews) or some elderly less comfortable with technology.


I've listed five ways our public transport's policy, planning and management has been anti-poor. Some are the result of long-ago infrastructure decisions, exacerbated by trends such as gentrification beyond a transport department's control. 

Others result from decisions about service priorities that could have easily gone a different and fairer, way. It's not even always a contest between good finances and social equity; sometimes we can have both. For example, while we generally have an underserviced network, there are cases where redistributing service from rich areas to a poor areas boosts both equity and fare revenue due to increased patronage. Similarly policies as disparate as the Free Tram Zone or airport rail impose substantial opportunity costs over alternatives such as a growth area rail extension or a Job Ready bus network. 

Certain polices on fares and practices on passenger information are also anti-poor. Fares policy is notoriously political. That disadvantages the poor because, at least in Melbourne, they are concentrated in safe seats. Passenger information may be more administrative with minimal accountability. Our Department of Transport appears less transparent than (say) American transit agencies. 

The activist landscape is also different. For example our PTUA committee is largely white and white collar versus LA's BRU being largely non-white and working class. Melbourne got train service improvements when commuter discontent electorally threatened governments. However we lack LA's popularly-based civil rights perspective. Activists in both cities challenge infrastructure projects such as freeways on environmental and social grounds but only the BRU has been able to use the threat of legal action (accusing 'transit racism' and a breach of the Civil Rights Act) to win fare cuts and  service gains through a 'consent decree'. Also, although Melbourne's ALP-affiliated RTBU and TWU  driver unions are industrially and politically powerful they are weak on service advocacy despite the jobs a boost would create. Our bus operators association, another beneficiary of increased services, has also been relatively quiet, unlike 10 to 15 years ago when they gained some large upgrades. 

Things may be less fraught here because unlike the US experience we've rarely seen big infrastructure being built as bus networks were being savagely cut. Instead we've had new infrastructure alongside service stagnation. Still, that could intensify pressures given population growth. Especially in lower income outer suburbs which are denser than their counterparts of 40 years ago yet are still being provided with similar infrequent service. Combined with forces such as gentrification and wage inequality I wouldn't discount the future possibility of LA-type grievances given the neglect of bus routes, including those in poor areas, that have less service than they did forty years ago. 

Always ask "who wins?"

One should ask 'who wins' out of any project to roughly gauge its equity implications. Avoid the temptation to pad pet projects with so-called 'wider benefits'. These are to a planner what a lamp post is to a drunk when projects don't stand up on transport grounds alone. Also think about opportunity costs and unintended consequences despite these being ruled out of scope.   

As an example freeways parallel to train lines can 'hollow-out' public transport patronage, make service improvements less likely and slow trams where they dump their traffic. Airport rail could be regressive since frequent fliers are overwhelmingly rich and many beneficiaries aren't even local taxpayers. Regional commuter rail taxes non-users to subsidise the exurban lifestyles of high income city workers paying insanely low myki pass fares. Level crossing removals may have wider benefits,  spreading well beyond the top end. Even so the cost of just one grade separation can fund $20k wombat crossings to improve pedestrian access at thousands of locations, several large roundabout removals or a decade of vastly improved bus connections to local homes and jobs. A project might seem good in isolation but it's always worth thinking about what else is good and possible. 


I've reviewed five ways that current public transport policies and services, even if not intended, could be anti-poor. There would not doubt be more. And some non-transport policies such as zoning, land use and housing. Comments are invited if you don't think some of the above are real issues or if there are others I might have skipped or which are more nuanced (eg policing and ticket inspections).

You might enjoy these well-regarded books on transport topics

Better Buses, Better Cities: How to Plan, Run, and Win the Fight for Effective Transit Steven Higashide 

The Public City: Essays in honour of Paul Mees Gleeson & Beza

A Political Economy of Access: Infrastructure, Networks, Cities, Institutions (Access Quintet Book 4) David Levinson

Human Transit: How Clearer Thinking about Public Transit Can Enrich Our Communities and Our Lives Jarrett Walker

Transport for Suburbia: Beyond the Automobile Age Paul Mees

(Sales links: I get a small commission if you buy via the above - no extra cost to you)


Andrew said...

Let's face it. The poor are constantly disadvantaged in all kinds of areas. Does build it and they will come, that is improve bus services in poorer areas, hold true? I would love poorer people in outer suburbs to have public transport like I have but we do have a terribly sprawling city.

Peter Parker said...

Yes it does hold true. There are many routes that had no Sunday service before minimum standards were introduced in 2006 - 2010 that now do well on Sundays. In both poorer and more affluent areas. Bus networks in areas like Point Cook, Tarneit and Craigieburn are carrying much more people than they were before services were upgraded. Some have needed further upgrades due to being so successful and continued population growth. Some stories about outer suburban sprawl are now out of date since density is now quite high in some growth areas (similar to some inner areas with trams).

Vicjenjen said...

Hahaha have a look at areas services by VLine. Kilmore is great to study, less than an hr into city but no buses in middle of day or any at all on weekends, no buses even trains late enough for people to work nights In the city and it a predominately white suburb.

Peter Parker said...

Kilmore's almost a basket case for efficient coverage. Terrible disconnected streets and station out of town. Not much can be done without a bridge (even if single lane big enough for a bus) at the end of Tootle St.