What workers are willing to pay for jobs at safer times of the day

Not all jobs pay the same. Some jobs are pleasant, fun, and/or clean. Other jobs are unpleasant, boring, and/or dirty. Because more people want to do the pleasant, fun, and clean jobs, the wages for those jobs tend to be lower (holding other characteristics of jobs the same). And because fewer people want to do the unpleasant, boring, and dirty jobs, the wages for those jobs tend to be higher (holding other characteristics of jobs the same). This is an explanation of what economists call compensating differentials. Jobs with desirable non-monetary characteristics tend to have lower wages than jobs with undesirable monetary characteristics. Workers are essentially compensated for taking on jobs with undesirable non-monetary characteristics.

One of the important non-monetary characteristics of a job is how safe or risky the job is. Jobs that are risky tend to pay more than jobs that are safe. Usually, economists think about this in terms of how safe or risky the job itself is. However, the same theory should apply to jobs in risky locations, compared with jobs in safer locations, or to jobs at risky times of the day, compared with jobs at safer times of the day.

How much are workers willing to sacrifice to work at a safer time of the day? That is the question that was addressed in this recent article by Oscar Becerra and José-Alberto Guerra (both Universidad de los Andes), published in the Journal of Economic Behavior and Organization (sorry, I don't see an ungated version online). They used an experimental approach, where they offered:

...participants a one-hour job and elicit individual preferences about alternative arrangements affecting the job’s safety perception by exploiting the time of the day (early versus late shift) and location (online versus on-site). In the first round, we offer students to participate in a second round in which they must perform an unspecified simple task at the university campus, which is located in a neighborhood generally perceived as more unsafe than most of the city. Participants are then allowed to choose between an early shift (9–10 a.m.) and a late shift (8–9 p.m.), given a compensation schedule of 11 fixed payments. We told participants, without deception, that we would invite them for this future one-time job under the proposed payment scheme. Based on their choices, we recover a measure of the willingness to pay for the early on-site shift. In the second round, a subsample of subjects from the first round faced a similar setting but now the future task could be completed remotely, which removes safety concerns related to the workplace’s neighborhood and commute. 

The research participants were all students at Universidad de los Andes in Colombia. Becerra and Guerra document that there are negative perceptions of safety around the university campus at night, and then based on their experiment they find that:

...safety concerns about the late shift on campus and gender differences (beyond those in safety concerns) are the main determinants of willingness to pay for the early on-site shift. The unconditional difference in willingness to pay between individuals who consider the late shift as riskier than the early shift is 3,721 Colombian Pesos (COP) –about USD 2.8 PPP and 11%–15% of the baseline payment. Similarly, we also find an unconditional gender gap in willingness to pay for the early on-site shift of 2,843 Colombian Pesos (COP) –USD 2.1 PPP and 8%–11% of the baseline payment. Using regression analysis, controlling for safety concerns reduces the gender gap by 20%.

So, workers are willing to sacrifice a relatively large proportion of pay in order to work in a job at a safer time of the day. Workers who feel less safe are willing to sacrifice more if they have high safety concerns, and women are willing to sacrifice more than men (even accounting for women having higher safety concerns than men). Becerra and Guerra then go on to show that there is likely to be external validity for their results, because based on administrative data from the university:

...students whose willingness to pay for the safer shift is larger than the median participant were more likely to exit University premises before 5 p.m., and they were also less likely to enroll in courses with lectures occurring after that time.

In other words, the experimental results were consistent with actual behaviour, in relation to being on campus in the evenings. So, as expected it appears that workers are willing to pay (11-15 percent of their wage) for jobs at safer times of the day. That means that there should be a substantial compensating differential for jobs that require working at unsafe times.

The gender difference in willingness-to-pay for working at safer times of the day has interesting implications. It means that, given a particular wage on offer, women are less likely to accept a job at an unsafe time (like evening or night work), so we should expect to see more men than women working at those times. It also means that women would be willing to accept a lower wage for working at safer times of the day than men do, since women are willing to give up more wages to work at those times than men. This could be an unrecognised contributor to the gender wage gap in locations where personal safety is generally low (or where risks are particularly high at certain times of the day). This is definitely something that is worth further exploring in other settings.