In part one, based on N.S.W. Roads and Traffic Authority accident data, overall trends in Sydney metropolitan taxi accidents are discussed, with results including analysis of accidents by severity, age, time of day, day of week, and month of the year.

The major findings indicate that taxi accidents do not differ markedly in severity from an appropriate general public comparison group, but that taxi accidents differ from the public in terms of both age and time measures. Taxi accident rates are most elevated at the end of weekend night shifts, suggesting a “black time” (Folkard, 1996) that results from the combined effects of long shift hours with the natural low point in the human circadian rhythm.

Concerning types of accidents, collisions with pedestrians are over-represented among taxi drivers, particularly accidents that result in fatalities and serious injury.

In part two, based on a survey of 165 Sydney taxi drivers, detailed analysis of the factors associated with accidents is presented. In addition, basic data concerning issues such as work patterns, attitudes, fatigue and personality are provided.

The major findings indicate that anger and risk-taking are important predictors of taxi driver accidents, with increased anger expression and increased risk-taking being related to a greater likelihood of involvement in accidents.

Average length of shifts and vehicle type were also significant predictors of accident involvement. (Comment: Where does driving style fit in ¬isnt is a significant predictor of crash rates? )

Once these four variables are taken into account, many other factors normally assumed to be responsible for taxi accidents, such as age, time holding a car or taxi license, kilometres travelled, employment type, shift type, etc. were not significant predictors of accidents. While risk-taking was a significant predictor of accidents, optimism concerning one’s driving abilities was found to be unrelated to both risk-taking and accidents.

Taxi drivers with sleeping problems were found to be much more likely to have fallen asleep at the wheel than other drivers, although overall rates of this occurrence are low. Aggression and sensation seeking were both found to be related to risk-taking, although economic pressures are suggested as an additional factor in taxi driver risk-taking.

Finally, taxi drivers work long hours per week (58 hours average total work), but do not seem to earn high levels of income.

In part three, based on feedback from surveyed taxi drivers and others, information concerning the nature of the job, together with discussion of problems within the taxi industry are presented.

On the whole, taxi drivers are profoundly negative about their working conditions and the structure of their industry. Particular problems include lack of driver safety, low earnings and lack of community respect. While these problems are specific to the experiences of Sydney taxi drivers, they illustrate systemic problems within the taxi industry that may be common elsewhere.

A paragraph or two on what the research suggests in terms of countermeasures and improvements within the taxi industry might go well here.

The most important theoretical development of this study is the finding concerning risk-taking and optimism about driving. Previous theories have assumed that individuals with an optimistic view of their driving abilities are likely to take more risks while driving due to over-confidence and perceived invulnerability.

This increased risk-taking ironically leads to increased accident rates due to over-estimation of driving skill. The current study found that while “optimism bias” is common among taxi drivers, it is unrelated to actual risk-taking while driving, and is also unrelated to accident rate. However, risk-taking itself (caused, in part, by aggression, sensation seeking and need for income) is a significant predictor of accident involvement.

Consideration is given to ways of reducing anger and risk-taking among taxi drivers, and speculation concerning the differential effects of optimism bias on both experienced (taxi drivers) and inexperienced (young) drivers is presented.