First published in Health and Safety at Work Magazine, February 2016. Illustrations for this version, plus updated references (Ahmed and Asim Faheem, 2021 and Oswald, Sherratt and Smith, 2017) added in 2021.
When the carrier bag charge was introduced in England in October 2015, bag use dropped. UK Government statistics suggest that by 2020 bag sales at major supermarkets had reduced by 95%. Telling people that carrier bags would damage the environment for our grandchildren was not sufficient incentive for people to change their habits. Charging them five pence was.
Similarly, though we might think that staying alive and coming home from work with fingers and senses intact is enough reason to behave safely, sometimes people need more motivation. “Do it this way because it’s safer” doesn’t work if people believe they can do it more efficiently and “safely enough” their own way. Unless there is the risk of getting caught by a speed camera, some people will judge it “safe enough” to drive at 40mph in a 20mph limit near a school, because it’s the holidays. In the workplace, staying safe is not sufficient motivation to “waste” time fetching a step ladder, when experience has shown that climbing on a chair will be fine.
There is a lot of literature on motivation theory but for simplicity we will consider two polarised viewpoints. The first was made popular by management theorists such as Frederick Taylor at the end of the 19th century. One element of Taylorism is that workers are motivated by money and will work harder if they know they will benefit from the success of their enterprise. Link bonuses to success and workers will succeed. Though some aspects of Taylorism were softened in the 20th century, the view that people are primarily motivated by money still persists in the minds of many.
By contrast, the work of psychologists such as Frederick Herzberg and Abraham Maslow during the middle of the 20th century presented more nuanced concepts of motivation. If someone does not have enough money for the essentials, they can be motivated to get money. Once someone has “sufficient” pay, however, they will not be motivated by small increments (in the form of bonuses), but rather by recognition, by the possibility of promotion or by intrinsic aspects of the task itself, aspects that lead not to a larger bank balance but to self-actualisation or, as Maslow wrote in 1943, “the desire to become everything one is capable of becoming”.
But if people are not motivated by bonuses, why do so many organisations have bonus schemes for other aspects of performance, such as meeting production or sales targets? And if 21st century organisations still believe workers can be motivated to work harder using bonuses, why do so few believe that they can be motivated to work more safely by using a similar mechanism? What message does that send?
One argument against safety-related incentives is that providing a bonus based on a target for reducing accidents will result in underreporting. In 2011, the Rail Safety and Standards Board (RSSB) reported that Network Rail employees and contractors had been massively underreporting accidents in order to maintain bonuses and, in the case of contract staff, to keep their work. RSSB estimated that up to 600 RIDDOR lost time injuries may not have been reported by between 2005 and 2010.
David Sutherland, now director of a risk management company in Australia, was an electrical apprentice in a food production factory in 1978, and his experience turned him against financial incentives. His employer introduced a scheme where there was a bonus, equivalent to 20% of his salary, once the factory had gone one month accident free. This payment was made every week as long as there were no accidents. It was a huge incentive but it backfired. “Someone had his thumb torn off in a machine,” he explains. “Some guys in the factory were so annoyed to lose their bonus they beat him unconscious.”
Gary Lindsay, health and safety manager at vocational training college Warrington Collegiate, maintains that sometimes rewarding the results can work. In the 1990s, he was employed in the spot welding department of a car factory where there were a lot of problems with people not wearing eye protection and suffering eye injuries from the sparks. The company introduced a scheme in which the group with the fewest accidents received an amount of money for their nominated charity.
“It didn’t result in underreporting but it did encourage people to wear their goggles and reduced the number of accidents massively,” he says.
One reason this results-based scheme worked, while the one David Sutherland experienced didn’t, could be the influence of individual behaviour on individual injury. It is fairly clear that if you are welding and you do not wear provided eye protection, it’s probably your own fault if you get injured. But if you lose your thumb in a machine, for example, there could be many people involved in a chain of procurement, maintenance, inspection and supervision failures.
Research within hazardous industries by Maslen and Hopkins (2014) published in the journal Safety Science supports this idea, finding: “most people did not perceive [injury rate measures] to be within their control”. In particular, where zero injury frequency rates had been set as a target, respondents regarded such targets as unachievable and therefore discounted the target altogether when they considered how to do their jobs.
Maslen and Hopkins also reported some more positive results. SMART – in this case, specific, measurable, assignable, realistic and time related – targets based on behaviours that could be “operationalised … into their daily work”, rather than results, were more successful in driving safe working. Positive examples included speed of dealing with equipment alarms, welding quality assurance and training.
Although Maslen and Hopkins found examples of performance agreements with up to 30 items, they concluded that setting more than five targets left people feeling that the goals were unmanageable. Research by Mattson et al in Human Resource Management Review (2014) at three Swedish nuclear power plants with different bonus schemes reflected this view that people respond better to a smaller number of clear, uncomplicated goals, than to complex bonus systems. They also identified a contradiction: though it might seem better to reward individuals for their specific behaviour, bonus schemes where everyone gets the same reward are considered fairer. This leaves employers planning incentive schemes with a dilemma: how do you motivate individual safety behaviour and satisfy staff that the system is fair?
Perhaps this is why some organisations provide recognition through donations to charities. Gary Lindsay’s eye protection example did not involve individuals benefiting from their actions – except in so much as they saved their eyesight – but relied on staff satisfaction that a charity had benefited. As reported in HSE research on Safety Culture on the Olympic Park (RR942), charitable donations were used as a motivational tool during the construction of the 2012 London Olympic Park. One contractor pledged a charitable donation of £1 for each (positive or negative) observation card submitted by its workers. This scheme eventually raised more than £6000.
Both the Maslen and Hopkins case studies and the Mattson et al research identified that often people were not motivated by the money itself, but that money was an important indicator of what mattered to an organisation.
Mattson et al stated that “money in general was largely considered motivational in that it reflected the employer’s appreciation”, while Maslen and Hopkins concluded “the bonus’s motivational impact turns on the feedback it provides to employees on how much their performance is valued”.
In another example from the 2012 Olympic Park build, a contractor rewarded the employee of the month – as nominated by their colleagues for their engagement in health and safety – with both a financial payment and a free breakfast with the project manager.
Across the project, other workers were offered rewards such as badges, clothing and lanyards, which had a low financial value but were highly visible to colleagues and managers. Other low cost incentives included vouchers for meals or the opportunity to go home early on a Friday afternoon. RR942 is not always clear exactly what the rewards were for but, where it provides examples, they are behaviour-based, covering actions such as stopping work if there was a safety problem or reporting unsafe conditions; they were not linked to accident numbers.
Money: intrinsic or symbolic?
“money in general was largely considered motivational in that it reflected the employer’s appreciation”
Mattson et al
“the bonus’s motivational impact turns on the feedback it provides to employees on how much their performance is valued”
Maslen and Hopkins
Financial incentives don’t work in a vacuum. If the safety team offers a bonus for stopping work if someone is concerned about safety, but the supervisor chides the worker for the delay, the immediate concerns of the supervisor are likely to win out over the safety incentive. The examples from the Olympic Park development worked because the overall safety culture was consistent: “work safely or not at all”.
When John O’Malley was health and safety coordinator for the Impress metal packaging manufacturing plant in Merthyr Tydfil, “Safety Bingo” was introduced as one element in a safety improvement scheme. The scheme also incorporated a safety message from the chief executive that everyone had to watch, increased machine checks and thorough accident investigations.
Each worker had a unique card of numbers and a new number was issued each day there were no accidents or infringements, such as not wearing personal protective equipment (PPE). When there was a winner, new cards were issued and the prize money went up for the next game. An accident voided the game and reduced the prize money for the next card.
Unlike David Sutherland’s example, Safety Bingo didn’t result in unhealthy competition because the safety culture was improving at the same time. “What we got was unexpected but much better,” explains John O’Malley. “The guys on individual lines pooled their cards, so when someone challenged the unsafe behaviour, the conversation was more like ‘Don’t do that, it’s not safe. We only need two numbers to win the bingo and if you have an accident we’ll lose the prize.’ That type of challenge proved very effective.” Lost time accidents were cut from 14 in 2004 to two a year in 2006 and 2007.
One of the failings identified by investigations into the 2005 BP Texas City oil refinery explosion in the US was that bonuses were linked to reported personal injury accidents, such as those resulting from slips and trips, and to the speed of turnaround. The first of these encouraged underreporting of what was in any case the wrong indicator, and the second encouraged process safety risks. An unsurprising recommendation from the 2007 independent Baker panel report, which looked at safety culture and management systems across BP North America in the aftermath of Texas City, was that bonuses should be linked to process safety indicators.
Jeremy Stewart, now health, safety and environment manager for a specialist electronics recycling contractor, Belmont Trading, previously worked for a large waste management company that introduced bonuses linked to safety. It avoided the problems associated with underreporting by linking the bonuses not to accident figures but to near miss reporting. Stewart recalls, however, that the system didn’t work. “Each employee was required to report a minimum number of near misses to qualify for the bonus,” he explains. “So you ended up looking for any old near miss to make up the numbers, rather than identifying critical events.”
One example was that some people reported vehicles speeding on the access road every month, “something we couldn’t influence, no matter how often it was reported,” says Stewart. He also worked with BP Grangemouth, where a different approach was more successful. There, staff received shopping vouchers based on the quality of near miss reports. “It was more meaningful to us,” he stresses, “and avoided time-wasting reporting just to meet a target.”
A more recent study provides a warning not to be blinded by short-term benefits from an incentive scheme. Having weeded out 42 organisations with a poor safety culture, researchers Ishfaq Ahmed and Asim Faheem were left with three organisations that were “good at safety practices.” In a paper in Safety at Health at Work (2021) the authors explain how incentives appeared to improve safety performance and motivation in the short term – but the effects didn’t last. After three months, performance and motivation were back to pre-incentive level, and comparable with the control group.
If you give a financial reward for increasing output, making more sales or reducing waste, is it really good enough to say that improving safety is sufficient reward in itself? The money itself might not motivate but without it the message received may be that the organisation values the sales and productivity more than safety or health.
The Baker report didn’t criticise BP for having a safety-related bonus system; it criticised it for giving bonuses for the wrong things. It’s easy to measure accident rates and near miss reports, but counting the first can result in underreporting, and the second can lead to spurious work creation. Identifying the behaviours that result in a safer workplace – such as carrying out risk assessments, providing supervision and responding to alarms – is the first step in devising a safety bonus scheme. Look out for the unintended consequences and do it all in the context of improving the culture that motivates safe behaviour.
An investigation into rewards on UK construction sites by David Oswald, Fred Sherratt and Simon D Smith (2017) provides more guidance about what types of incentives are welcomed, and some of the pitfalls. For example, being hailed as a safety winner could be seen by some as a stigma, rather than a positive. Consistency is also important to people – starting with high value rewards, and dropping to lower value prizes will leave people frustrated rather than motivated.
A final warning comes from a meta-analysis by Weibel et al (2009) published in the Journal of Public Administration Research and Theory. Though not specifically looking at bonuses for safety performance, this found evidence that while bonuses can be effective for uninteresting and simple tasks (which would include wearing PPE), they can have a detrimental effect on interesting and difficult tasks, such as the problem solving that would be necessary in an emergency.