Table of Contents
A survey of cannabis use by programmers reveals that over a third had tried programming while using cannabis and almost one in five do so at least once a month, motivated by perceived enhancement to their ability to get into a programming zone.
This is despite the fact that anti-cannabis policies and screening remain common for programming-related jobs, with 29% of respondents reporting having taken a drug test for a programming-related job.
Cannabis is the source of much controversy. It has widely acknowledged benefit for pain relief and is a popular recreational drug for its mind-altering properties. However, being classed as a Schedule 1 drug (which defines it as having no therapeutic value and a high abuse potential) its use in much of the USA is illegal, with only 17 states having legalized it. On the other hand, 91% of Americans believe cannabis should be legal for medical or recreational purposes and 81% of Americans believe cannabis has at least one benefit, mostly medical – and 36 US states allow the use of medical cannabis.
Until now there have been no empirical studies of how cannabis interacts with software development, a gap filled by a survey conducted by a team of three researchers at the University of Michigan, Ann Arbor that will be aired at the 44th International Conference on Software Engineering (ICSE 2022), scheduled to take place in Pittsburgh, May 21-29, 2022.
Official statistics indicate it cannabis is used by 17.5% of Americans aged 12 or older but its use among the software developer community is disproportionately high. To quote from the arXiv paper, Hashing It Out: A Survey of Programmers’ Cannabis Usage, Perception, and Motivation:
Cannabis use in Silicon Valley appears to be high, with area dispensaries reporting that around 40% of their clientele are tech workers. Similarly, a qualitative study of coding bootcamps identified “lots and lots of [weed]” as one key element of support
To investigate cannabis prevalence, usage motivations, and culture in software engineering, Madeline Endres, Kevin Boehnke, and Westley Weimer conducted an online survey of 803 US-based programmers, recruiting participants through GitHub and social media as well as at the University of Michigan.
The demographics of the survey population, broken down by recruitment pool is in this table:
Note: While those in the ‘University’ column are those recruited from university emails many had already graduated and thus had fulltime CS jobs. Multiple options were permitted for Employment Status and Job and many selected both Software Engineer and Developer.
Of the 803 people surveyed 450 had full time Computer Science jobs, the majority of which can be considered progamming roles. The vast majority were in the age range 20-35 and men outnumbered women 6:1.
The research addressed four questions and here are its findings in brief:
Q1. Do programmers use cannabis while programming? If so, how often?
Overall, we find that 35% (280/803) of our participants have tried cannabis while programming or completing another software engineering-related task, approximately half of those who tried cannabis in general (69% = 557/803). Of those that have used cannabis while programming, 73% (205/280, 26% of our population overall) used cannabis while programming in the last year.
In the last year, 53% (147/280, 18% of our full sample) reported using cannabis while programming at least 12 times (monthly). Furthermore, 27% (76/280) reported using while programming at least twice a week (100 times per year) and 11% (30/280, 4% of our sample as a whole) reported using on a near daily basis.
The survey also revealed that cannabis use while programming is particularly common for non-binary or transgender participants (57%) and participants recruited from GitHub (39%).
Q2. In what contexts do programmers use cannabis?
Cannabis tended to be used more for personal and non-urgent programming tasks. The figures here relate to the 280 respondents who used cannabis while programming:
Cannabis use was also more prevalent when working remotely:
We find that 52% (145/280) of cannabis-using programmers report they are somewhat or a lot more likely to use cannabis for work-related tasks when working from home compared to only 5% (13/280) who report that they are somewhat or a lot less likely to do so.
Asked about the type of programming activity they would use cannabis for programmers reported a higher likelihood of using cannabis while brainstorming or prototyping and a lower likelihood of using cannabis while performing quality assurance, requirements elicitation, or tasks with an imminent deadline. In the chart it is the blue part of the line that indicated increased likelihood of cannabis use:
Q3. Why do programmers use cannabis?
Overall, we found that programmers were more likely to report enjoyment or programming enhancement motivations than wellness motivations: the most common reasons were “to make programming-related tasks more enjoyable” (61%) and “to think of more creative programming solutions” (53%). In fact, all programming enhancement reasons were selected by at least 30% of respondents. On the other hand, general wellness related reasons (such as mitigating pain and anxiety) were all cited by less than 30% of respondents. Thus, while wellness does motivate some cannabis use while programming, it is not the most common motivation.
Q4. How do opinions of programming cannabis use vary between managers, employees, and students?
Regarding the perception of cannabis among those surveyed the paper states:
Programmers in our sample have more positive attitudes towards cannabis than the population overall. For example, 91% of our participants say that marijuana use should be legal for both recreational and medicinal use compared to 60% of the general United States population in 2021. Similarly, only 5% of our population views smoking marijuana once or twice a week to be of “great risk” as opposed to 29% of the US population. This difference is likely explained by the demographic differences in age, gender, and political leaning between programmers in our sample and the population overall.
Having found no significant differences in cannabis usage between mangers, employees and students there were some differences in perceptions between the three groups. For example, the researchers found that managers disapprove of cannabis less than employees expect them to!
Hashing It Out: A Survey of Programmers’ Cannabis Usage, Perception, and Motivation
by Madeline Endres, Kevin Boehnke, and Westley Weimer
In Proceedings of The 44th International Conference on Software Engineering (ICSE 2022)
Happy Programmers Produce Better Apps
On the Unhappiness of Software Developers
Creator of arXiv Wins Einstein Award
Pair Programming – Principle and Practice
or email your comment to: [email protected]