Menu Close

FactCheck: do poor people drive less?

Rich or poor, being stuck in traffic is always annoying. Shultz6/Wikimedia Commons, CC BY-SA

“The people that actually pay the most are higher income people, with an increase in fuel excise… The poorest people either don’t have cars or actually don’t drive very far in many cases.” – Treasurer Joe Hockey, ABC Radio, August 13

Do rich people pay more?

The first proposition here is that the people who pay the most fuel excise are those on higher incomes.

This statement can be interpreted in different ways. Excise is not income-related, so higher-income households do not pay more tax per litre for their fuel. But in terms of absolute expenditure, data show that higher-income households spend more, on average, on petrol than lower-income households.

Australian Bureau of Statistics (ABS) Household Expenditure Survey Data from 2010 reports that households in the lowest 20% of income spent an average of A$16.36 on petrol per week, whereas those in the highest 20% spent an average of A$53.87.

But those figures relate to households, whereas the Treasurer referred to people, so the household survey data aren’t capable of supporting or refuting the Treasurer’s claim if we take his statement literally. A per-person calculation of fuel excise would be very complex, and we probably don’t have data sets that are detailed enough to answer this question.

The other factor is the degree of discretion in spending. Wealthier households are, all else being equal, more likely to use their car for discretionary rather than essential trips.

Do poor people own fewer cars?

The Treasurer’s second main proposition is that the poorest people don’t have cars.

This statement is verifiable against motor vehicle ownership data, the best source being the ABS Census. The Minister’s Office has provided ABS data on car ownership relative to the Socio-Economic Indexes for Areas (SIEFA), which measures relative household disadvantage.

Car ownership broken down by relative household advantage. Red portions indicate households with no cars. Treasury estimates based on ABS 2011 Census of Population and Housing tables from ABS 2033.0 Socio-economic Indexes for Areas (SEIFA) and

Assuming these data are accurate, they show that roughly 65% of households in the lowest 10% on the SEIFA index own cars, in contrast to 83% of households in the highest 10%. Lower-income households thus own fewer cars, on average, than higher income-households. But note that even among these poorest 10% of households, a majority nonetheless own at least one car.

Furthermore, SEIFA primarily measures income and indicators of disadvantage, rather than “wealth”. A household or individual may have low income but not necessarily be poor if they own large assets. So SEIFA is not an ideal data set to directly compare poor and wealthy households. In a Melbourne study that considered “forced” car ownership among low-income households in Melbourne, the data showed that 72% of low-income households own at least one motor vehicle.

Finally, car ownership does not equate to use, and is thus only a very indirect indicator of fuel cost.

Do poor people drive shorter distances?

The Treasurer went on to claim that poor people “don’t drive very far in many cases”. This is much harder to ascertain, as there is no survey that measures national travel patterns. It is also not clear whether these shorter distances are defined as overall kilometres, or kilometres per trip undertaken.

Metropolitan household travel survey data are not generally provided by state transport departments in terms of income categories, if they collect such data at all. Moreover, there have been few recent Australian studies that have appraised this relationship using such information. The Melbourne survey cited above also noted that “low-income households make less trips than equivalent middle- and higher-income households”, but that “travel distance per trip generally increases with the number of cars owned per household, regardless of income group”.

Plenty of studies have also linked lower-income households with a range of disadvantageous travel circumstances, mainly because of the geographical locations of suburbs where poorer people tend to live.

In Sydney, for example, people in areas with higher proportions of lower-income households generally make more and longer trips than those in more affluent areas. Also, the bias towards cars over other forms of transport tends to be greater in less affluent areas.

As this Brisbane study shows, people living in areas with lower average incomes tend to drive more and further (principally for work), own older and less efficient cars, and face relatively higher costs for operating motor vehicles.

In terms of options to offset higher fuel costs by switching to other forms of transport, the distribution of high-quality public transport typically favours wealthier households within Australian cities.

A survey of the energy requirements of Sydney households showed that lower-income households have very high automobile energy use due to their high car dependence, but that higher-income households generate greater carbon emissions, because of higher overall consumption. In this context the government’s shift in energy taxation from carbon emissions to taxing petrol use seems to be socially regressive – many less wealthy motorists have no option but to take the hit.

In summary

So, what to make of the Treasurer’s statement? The terms he used are loosely defined, which makes them difficult to assess against the data. At a basic level, Mr Hockey didn’t say where the threshold is when considering who are “the poorest” people.

At a national scale his statement contains some verifiable truth. But at the scale of regions or cities, spatial factors such as the suburbs where people live and the transport links they use are much more important.

Finally, it is better to think of fuel costs in proportion to disposable income, rather than as absolute costs. We should also consider people’s access to alternative options such as buying a more fuel-efficient vehicle, or switching to public transport, walking or cycling. People with high discretionary income or who live in areas well served by public transport (and there is a significant overlap between these groups) will obviously have more options open to them, whereas poorer households who lack access to public transport might simply have to cut back on travel.


“Higher-income people pay the most fuel excise”: conditionally true in absolute terms and as a general category, but not necessarily true as a proportion of income or of discretionary spending.

“Poorer people either don’t own cars or don’t drive very far”: unverifiable in any meaningful way. Car ownership is marginally lower in poorer households, but is nevertheless high right across the board, with reason to believe many poorer people are forced into driving long distances (mainly for work) by their geographical circumstances.


This is a good breakdown of the Treasurer’s statement and provides evidence to both verify and refute it.

It would be useful to describe the characteristics of the geographical locations that underpin the relationship between low income and disadvantageous travel circumstances. For example, why do people living in areas with lower average incomes make more and longer trips by car? Is it because these areas are traditionally in the middle to outer suburban ring and are comparatively poorly served by public transport and employment opportunities and services?

Another aspect to this issue is the socio-economic gradient across the rural-regional-urban divide. Much of the focus has been on cities, but low incomes and car dependence are also important issues in rural and regional areas. – Jennifer Kent

Have you ever seen a “fact” that doesn’t look quite right? The Conversation’s FactCheck asks academic experts to test claims and see how true they really are. We then ask a second academic to review an anonymous copy of the article. You can request a check at Please include the statement you would like us to check, the date it was made, and a link if possible.

Want to write?

Write an article and join a growing community of more than 152,900 academics and researchers from 4,487 institutions.

Register now