A possible scenario for Sydney: TRIPS ON MOTORISED MODES BY RESIDENTS ON A TYPICAL WEEKDAY IN 2050 (source data: Transport for NSW)

I’ve previously noted that the share of residents’ travel captured by public transport in Australia’s capital cities has barely budged over the last 40 years (see Where to with transport in our capital cities?). Nor has it improved much in recent years; for example, cars’ share of motorised travel in Sydney was 86% in 2006-07 and 85% in 2016-17. In Melbourne, it went from 91% to 90% over the same ten-year period.

But what if public transport were to enjoy vigorous, sustained growth for an extended period, say from now until 2050? What impact would it have on mode share if it were to grow much faster than car use? Critically, what scale of infrastructure investment would be required to support the additional trips?

A possible scenario

The exhibit sets out a possible scenario for residents of the Sydney metropolitan area. I’ve used data on weekday trips by mode from the household travel survey run continuously by Transport for NSW since 1997 (trips is a more useful measure than kms of travel for this discussion because it impacts directly on infrastructure capacity).

I’ve created a scenario for the 32 years from 2017/18 to 2049/50. I’ve assumed:

  • Total motorised trips (by car, rail and bus) grow by 54%, roughly in line with projected population growth of 51% over the period.
  • Rail trips increase at a compound rate of 5% p.a., considerably higher than the rate of growth experienced over the last seven years (3.9% p.a. compound).
  • Bus trips increase by 1% p.a compound, a little lower than the growth rate over 2010/11 – 2017/18 (1.3% p.a. compound). The assumption recognises a large proportion of the growth in rail travellers will come at the expense of buses.
  • Car trips increase by 0.6% compound, significantly less than the 1.4% p.a. rate they grew at over the last seven years. This is considerably slower than projected population growth and is consistent with the evidence that car use is already close to ‘saturation’ levels.


Thus I assume public transport trips (rail + bus) are projected to grow at 3.7% p.a. compound to 2050, compared to 3% compound over the last seven years. This is a more ambitious assumption than it might at first appear, since population grew at a much higher rate over 2010/11 – 2017/18 (4.4% p.a. compound) than it’s projected to over the next 32 years (1.3% p.a. compound).

I don’t consider ferries because I don’t have comparable data. Their share of total trips is in any event small at present e.g. 0.3% of AM peak trips.

Growth in trips

Given these assumptions, total motorised trips by residents on a typical weekday increase by 8.3 million over the 32-year period to 2050. By mode:

  • Rail trips increase by 271%, from 1.2 million to 5.7 million.
  • Bus trips increase by 37%, from 1.0 million to 1.4 million.
  • Car trips increase by 37%, from 13.1 million to 15.8 million.


Change in mode share

The exhibit shows the resulting change in (motorised) mode share from 2017/18 to 2049/50

  • The share carried by car falls from 86% to 69%, although the number of car trips increases by 4.9 million.
  • The mode share of public transport (rail and bus) more than doubles, rising from 14% to 31%, with the number of weekday trips increasing by 3.4 million.


Implication: massive increase in rail capacity

Achieving a near-tripling in the number of Sydneysiders travelling by rail in 2050 would require an enormous commitment to new infrastructure. While I expect an increasing proportion of travel would be off-peak, the majority would still be in the AM and PM peaks. In order to handle the much larger peak loads, it would be necessary to build new lines as well as upgrade existing lines.

On top of that, the public transport system would likely still have to carry growing numbers of visitors to Sydney (since they’re not residents, visitors aren’t counted in the household survey). International and domestic visitors spent 110 million nights in Sydney in the year ending September 2018, up 27% over the last four years, adding significantly to the public transport task.

I’m not aware of any analysis of the specific infrastructure improvements required to handle growth on this scale, but we can reasonably assume they would cost hundreds of billions of dollars. While this level of investment would likely deliver benefits in excess of the costs, other opportunities for spending the money would necessarily be foregone (health and education are the big areas in the state budget).

Implication: focus on ‘taming’ cars

I think the most telling message from this exercise, though, is that even with strong assumptions about growth in public transport demand, the car still remains by far the dominant mode for travel by residents, accounting for over two-thirds of motorised trips (69%) in 2050. This does not suggest the end of the car is nigh.

The continuing ascendancy of the car underscores the critical importance of focusing on ways to ‘tame’ travel by private vehicle. Relying solely or primarily on building new rail infrastructure to reduce the environmental, amenity and congestion impacts of motoring is way too slow and simply nowhere near enough.

As I’ve noted before, concerted action is needed to make cars a lot cleaner, smaller, slower, safer, quieter and more considerate of other users of the city than has been the case to date. By itself, better public transport isn’t going to banish cars from Sydney’s roads (see Where to with transport in our capital cities?).

A car-taming scenario

Given the sheer scale of car use at present, the biggest gains would come from focusing directly on reducing the number of car trips.

For example, consider a second scenario where the growth in bus and rail travel over the 32-year period is as per scenario one, but a brace of policies is implemented that directly limits growth in car trips to zero.

This is a small difference – from the 0.6% p.a. compound assumed in the first scenario, to zero per annum – but it would nevertheless be very difficult politically. Possible actions include road pricing, higher parking charges and increased taxes on car ownership and operation.

This scenario reduces car trips in 2050 by 2.8 million each weekday relative to scenario one. If all of these “foregone” trips were replaced by rail, this small reduction would have a dramatic effect; car’s share of motorised trips would fall from 86% at present to 61% in 2050, while public transport’s would rise from 14% to 39%.

It’d require even more investment in infrastructure, of course, because instead of nearly tripling as in scenario one, rail trips would increase almost five-fold. On the other hand, there’d be savings from lower demand for additional road infrastructure.

Under a more ambitious scenario, public transport would (just) be the majority motorised mode (51%) by 2050 if car trips could be reduced at the rate of -0.5% p.a. compound over the period and all of the “foregone” trips replaced by rail. The number of rail trips in 2050 would be almost nine times higher than at present.

Change on this scale won’t happen just by building more rail infrastructure or providing more buses; the big pay-off will only come from reducing the competitiveness of motoring.

It should be noted that since car travel is currently under-priced, it’s likely some of these “forgone” car trips would be suppressed entirely rather than replaced by public transport. And given that the average length of car trips is short, some would be replaced by active transport i.e. walking and cycling. In other words, not all or perhaps even a majority, of the “foregone” car trips would be replaced by public transport.

The upshot

Let me emphasise that these are scenarios, not predictions. They nevertheless indicate that cars will remain by far the dominant mode even with fairly bullish projections for public transport growth. The clear implication is that cars aren’t likely to go away; much more attention must be given to ‘taming’ them.