Banner Image #1

Work health and safety of food delivery workers in the gig economy report

Read the full report here.

Executive Summary

Background and method

Food delivery workers (FDWs) in the platform-based gig economy are at risk of illness or injury while working. While the number of incidents involving FDWs that have been reported to SafeWork NSW is low, there has been an exponential increase over the past three years, from one incident reported in 2017 to 19 in the first half of 2020 alone. The majority of these incidents are road and traffic related, but previous research has highlighted additional harms that FDWs risk on the job, including physical assault, intimidation, and verbal abuse.

The bulk of research into the work health and safety (WHS) of those engaged in insecure work has, to date, focused on people who are self-employed, work casually, or are on temporary contracts. While the trends identified among these populations may be broadly applicable to FDWs in the gig economy, it is unknown to what extent this is true, or whether FDWs face unique WHS risks on the job. The global COVID-19 pandemic poses an additional and complex challenge to WHS in the gig economy. FDWs cannot work from home and in some cases are putting their own health at risk by delivering to those in self-isolation. As more inexperienced workers become FDWs in response to a widespread rise in unemployment, the risk of WHS harms may increase. Focused research into the WHS of FDWs is therefore needed to address this critical gap in both research and practice.

The Behavioural Insights Team, in collaboration with Macquarie University and the Centre for Work Health and Safety, is undertaking a four-phase project that aims to improve the WHS of FDWs in the gig economy. The objective of Phase 1 was to describe the characteristics of FDWs and their WHS behaviours, knowledge, and concerns. Phase 2 focuses on exploring the WHS behaviours, knowledge and concerns of app-based food delivery platforms (FDPs) who engage FDWs. In Phase 3, the findings of Phases 1 and 2 will inform the development of proactive, risk prevention activities to improve FDWs’ WHS, which will be co-designed with relevant stakeholders. In Phase 4, the outcomes of these prevention activities will be evaluated in the field.

This report details the results of Phase 1, in which we undertook five concurrent streams of data collection:

  • An online survey of FDWs
  • Semi-structured interviews with FDWs
  • A textual analysis of WHS-related posts on social media
  • A service safari, in which members of the project team signed up to the major FDPs, went through onboarding, and worked a shift as a FDW
  • Field observations of FDWs in restaurant hotspots

The data from each stream were then synthesised into the key insights, ideas and opportunities that will inform the subsequent phases of the project, in which proactive, risk prevention activities to improve the WHS of FDWs will be co-designed with stakeholders and trialled in the field.


Characteristics of FDWs

FDWs as a group are typically under 30, male, and in Australia on student visas, who choose food delivery work for its flexibility and to supplement existing income streams. While the existing evidence for the physical and mental health effects of gig work is mixed, we found that our survey respondents and interview participants reported better physical health than, and comparable mental health to, the Australian population as a whole.

FDWs tend to cluster in densely populated areas and have typically been in their job for less than a year, which accords with the preponderance of student visa holders, as well as a reported influx of workers due to COVID-19 (e.g. Keoghan, 2020). However, there is a significant minority of FDWs working in suburban areas and regional centres. The most common mode of transport differs on this basis, with urban FDWs more likely to use bicycles and e-bikes, and regional FDWs more likely to use cars and motorcycles. This suggests that these two populations may face slightly different WHS risks on the road and may thus require different and more tailored approaches to improving their road safety.

Health and safety behaviours of FDWs

FDWs engage in a range of behaviours that influence their WHS risk. Given that the overarching focus of this project is on developing proactive, risk prevention activities, we focused primarily on identifying the behaviours that FDWs engage in that both exacerbate and mitigate risk (i.e. proactive behaviours). Common risk exacerbation behaviours include using mobile phones while riding or driving, working while fatigued, wearing dark clothing at night, cycling on footpaths and in other pedestrian-only areas, failing to follow COVID-safe guidelines on social distancing and mask-wearing, and speeding or rushing. These behaviours are driven largely by the desire to work quickly and to maximise income. Common risk mitigation behaviours include riding or driving more cautiously and wearing brightly coloured or reflective clothing during inclement weather, wearing helmets, and having delivery vehicles serviced regularly. However, we also identified aspects of risk mitigation behaviours that pose risks in and of themselves, most notably FDWs’ frequent use of second-hand parts to self-service and repair their own vehicles.

There was a sizeable gap between self-reported risk exacerbation and risk mitigation behaviour (identified in the survey) and actual behaviour (observed in the field). This highlights the need to make WHS risks more salient and to encourage FDWs to think about and act upon these risks more consciously.

Health and safety knowledge of FDWs

Although FDWs believe that they are primarily responsible for managing their own WHS risks, the knowledge necessary for successful risk management tends to be acquired in a piecemeal fashion through a combination of the limited (and often optional) WHS information provided by FDPs, experience on the job, and advice from peers. Each of these learning channels has its own drawbacks. FDP information is most likely to be accurate and appropriately vetted, but the fact that it is typically not compulsory means that engagement with this information is likely to be low. The need for experiential WHS learning to reach a baseline level of risk awareness means that new, inexperienced FDWs are at greater risk of causing or incurring a WHS incident, and peer discussion forums can become fertile ground for misleading and even dangerous advice to flourish and spread. Overall, FDWs have a relatively low level of WHS knowledge, particularly in relation to Australian road rules and bicycle safety. This is likely due, at least in part, to demographic factors and the temporary, short-term tenure of most FDWs. Since knowing about WHS risks is a critical first step in taking appropriate action to reduce the likelihood of a WHS incident, improving WHS knowledge among FDWs is of key importance in improving overall WHS in this population.

Health and safety concerns of FDWs

The main WHS concerns we identified were traffic accidents, verbal abuse and robbery or vandalism of delivery equipment. We also found that WHS concerns and WHS experiences do not necessarily align: several of the WHS incidents about which FDWs are most concerned (e.g. traffic accidents, robbery) are among the most rarely experienced. This is likely attributable to the fact that high-impact, low-frequency events, such as a serious traffic accident or assault, tend to be more salient and thus more memorable than low-impact, high-frequency events. For the same reason, high-impact, low-frequency events are also more likely to become the subject of social media posts, which serves to both distort their likelihood and further magnify their impact on the FDW community. Verbal abuse, however, is unique in this regard: it is both frequently experienced by FDWs and ranks highly as a WHS concern, making it an ideal potential target for a proactive, risk prevention activity.

Opportunities and Ideas

Based on the insights of the research and a review of successful interventions from other policy areas, we identified eight ideas to take forward to the co-design and trial phases of this project. The ideas are aimed specifically at changing behaviour in order to improve WHS and are all underpinned by behavioural science. Each idea was rated in terms of its impact and feasibility. Impact was estimated by assessing the approximate number of people the idea would directly affect and the extent to which the idea would directly or indirectly improve WHS. Feasibility was estimated by identifying the range of stakeholders who would need to “buy in” to the idea, assessing potential obstacles, and estimating costs.

We note that the ideas represent opportunities only, rather than firm recommendations. That is, the focus of the co-design and trial phases of this project is not limited to the ideas we have presented in this report, which are summarised in the following table.

Table 1: Summary of ideas for risk prevention activities.


Risks addressed



Increasing participation in safety training at onboarding through making it mandatory or using better choice architecture

Lack of familiarity with Australian road rules and understanding of WHS obligations



Alter the order acceptance process to minimise the risk of phone distraction

Distractions on the road



Benchmark the time that FDPs are allowed to give FDWs for each delivery using map APIs

Rushing behaviours; stress due to perceived threat of robbery



Reduce abuse by humanising FDWs to potential abusers

Abuse from businesses, customers, and members of the public



Use a rating system to alter restaurant behaviour

Abuse from businesses, customers, and members of the public; COVID-unsafe waiting behaviours; stress due to perceived threat of robbery



Facilitate mentor schemes or buddy systems

Lack of familiarity with location-specific WHS risks



Send targeted and triaged safety prompts and reminders at critical points

COVID-19 risks; weather-related risks; abuse



Prompt FDWs to anonymously report near-miss and abuse hotspots and map them

Abuse from businesses, customers, and members of the public; environmental risks like poorly lit roads