New data drawn from the ongoing American Community Survey on what mode commuters use to get to work was released this week. Wendell Cox has summarised the numbers for the period from 2000 to 2011 (see exhibit).

The data indicates commuting habits in the US are changing at a glacial pace. Notwithstanding massive expenditure on new transit systems in many cities, driving still dominates the journey to work.

This data isn’t available for Australia yet, but we can learn a lot from North America. The usual caveats apply. Note also this data is only for the journey to work. It measures mode share by number of trips, not kilometres of travel.

The proportion of US commuters who drive fell 1.8 percentage points – from 87.9% to 86.1% – over the eleven years from 2000 to 2011, or by an average of just 0.16%  per year (from hereon, all percentages are ‘points’).  Worse, in the most recent year, 2011, the increase was much higher (0.76%), suggesting driving grew faster as the economy emerged from recession.

Notwithstanding the small annual decline in share, more Americans are driving to work. The absolute number of commutes by car increased 6.5 million over the period. In 2011 an extra 900,000 workers took to the roads.

Where did driving’s “lost share” go? Not a lot went to public transport. It increased its mode share from 4.6% to just 5% over the period, or by an average of 0.04% per annum (even lower, 0.03%, in the latest year). Patronage went up by a modest 1.1 million over the eleven years, considerably less than the increase in driving.

Cycling went from 0.4% to 0.6% over the period – impressive, but obviously from a small base. But the most significant  improver was home-based workers, who increased their share of all commutes from 3.3% to 4.3%. Their numbers rose from 4.2 to 6.0 million i.e. by 1.8 million.

That’s significantly larger than the increase in transit commuters. A lot of it is probably a direct result of the recession, but nevertheless the comparison highlights the relatively small gains made by public transport.

Those small advances are in spite of massive investment in rail-based systems in US cities in recent years, especially over the last decade.

For example, there are now 33 cities in the US with light rail or streetcar systems either completed or under construction. There are a further 42 US cities proposing to build systems (excludes heritage streetcar systems e.g.  San Francisco cable car).

Building a large market for public transport is extraordinarily difficult in a car-based culture. A key issue though is whether the hundreds of billions being spent on new rail-based systems represent the best way of winning travellers away from cars.

Back in 2006, Professor Peter Gordon from USC observed that “the $7.6 billion of rail added to LA County’s transit system over the last 15 years serves between one-quarter and one-half percent of all daily County trips.”

I noted something similar recently in the Australian context. The feasibility study for Melbourne’s proposed Rowville rail line, likely to cost in the order of $2 billion plus to construct, forecasts it would increase the share of all trips carried by public transport in Melbourne in 2046 from 12.6% to 12.7%, i.e. by 0.1%!

A big problem with public transport investment in the US is that the choice of projects is driven much more by political considerations than by good decision-making. As happens in Australia too, people are captured by the glamour and presumed worthiness of rail-based projects over more humble alternatives such as buses, or more difficult options like road pricing.

In a recent report, Game changers: economic reform priorities for Australia, the Grattan Institute noted the low benefit-cost ratios for many proposed transport projects in Australia. It sounded a warning about poor project selection:

Even if infrastructure is productive on average, it is only economically productive if it is “the right infrastructure, in the right place at the time and accessible at sensible prices”. Simply spending money on infrastructure is not enough to get a return if the cost benefit case is not there.

It will be extraordinarily hard to shift significant numbers of US (and Australian) workers out of their cars and into more sustainable modes. That however makes it all the more important to develop policies that deliver real mode shift.

As well as targeting spending on productive, cost-effective infrastructure, those policies should involve “soft” initiatives addressing, for example, issues like how travel is priced, how modes connect operationally to form a system, and how land uses reinforce and support desirable transport choices.

The numbers from the Community Survey also highlight the long time frames likely to be involved in significantly changing travel behaviour. Cars will very likely continue to be the dominant mode for a long time yet – both in Australia and the US – so they need to give much more towards the objective of a more sustainable, equitable and efficient transport system.

Of course it would be better to look at Australian data on commuting. The ABS won’t release 2011 Census data on the journey to work in Australia until the end of next month, so we have to wait.