2019 UtahJS Survey Results
In October of 2019, a survey of UtahJS members was conducted through the group's slack channel around the topics of pay and education. Presented here are an analysis of the results, as well as some commentary, and a link to the raw result data.
- Previous years
- Caveats & Biases
- Link to data & licenses
Tip: For charts with lots of data, hover over (or click) a chart series in the label to see those points highlighted.
I would consider myself a...
This year members were asked to bucket themselves in to job-titles, such as "Junior Developer" or "HR Manager"
Also, included in the Other category were (with 1 response each):
- Content Management/Media Design
- Principal Engineer
- Senior and Manager and Lead
- Software Architect Extra-Ordinair!
How much experience do people have in different roles?
This box plot shows similar results to previous years, with strong overlap of titles -- especially mid and senior level developers. Some observations:
- While the average for "Senior Developer" is 10 years, people are describing themselves with that label with as few as 3 years of experience.
- Junior developer is the narrowest band, with a range of 3 years from least to most experienced.
What are the experience levels for all of the survey respondents?
Gender of respondents...
This year the survey asked about gender, with the option of entering their own response in free-form text.
The responses for "other" are:
- Prefer not to say (1)
- Non-binary (1)
- Non-binary; genderqueer (1)
- Motorboat (1)
- Helicoptor (1)
- Apache Helicotor (1)
Editorial note: There were a few survey responses that I'm guessing were trying to make a joke, or some kind of statement about gender with their responses? Gender identity is such a sensitive and personal issue, that this sort of behavior puts me in an uncomfortable position of either 1) removing or changing someones gender identity response ("They survey editor deleted my gender preference!" I can hear them saying), or 2) including non-serious responses next to legitimate and personal "other" responses. I've chosen to include them, and it's my hope that their inclusion doesn't diminish or invalidate any of the other serious responses.
Gender comparison to 2018
This year, the survey included compensation questions about base pay (cash compensation), liquid equity (stock), non-reoccurring bonuses, pay changes from changing employers, as well as important benefits.
Two box plots are shown here: one including responses with zero compensation (students, out of work, taking time off), and one including those with jobs currently.
How does total compensation change with experience?
Of those that received equity or bonuses, what were they?
78 of 208 respondents reported receiving equity or a bonus.
Of those that did receive equity or a cash bonus, what was it?
How does gender relate to compensation?
Showing salary only
Zooming in on the experience area shows the distribution more clearly:
Does the distribution change when equity and bonuses are considered?
Zooming in on the experience area shows the distribution more clearly:
How does job title relate to compensation?
What is the shape of compensation growth over time?
Compensation growth over a career usually described along the lines of "lots of growth in the first few years, then it slows down once you reach the local market ceiling." I'd like to propose a different word to describe the shape of compensation growth: Logarithmic.
Usually, a logarithmic scale is applied to the Y axis when visuallizing a large range of values with high growth. We can, also however apply a logarithmic scale to the X axis (time in this case) to see how compensation growth appears to be linear when visuallized this way:
It is my observation that the strongest predictive variable in compensation seems to be a log function of years of experience, followed by (in no particular order) the industry type, technology/specialization, negotiating skills, and of course, an individuals relative talent.
Is compensation increasing from year-to-year?
Previous year surveys only asked about base salary, so we are comparing that metric below (not total compensation including equity or bonuses)
Tip: Hover over the year labels in the legend to see isolated year data.
- The responses between 2018 and 2019 do not wildly differ. This makes sense as you may not see a change in only 1 years time with the sample size that we have
- Select 2017 and observe that there are no points where a 2017 salary is the highest in it's respective experience range.
What is the distribution of education types?
Survey respondents selected from a pre-defined list of types of education, or "other" with free-form text
Other responses inlcuded different certificate programs, high-school only, and mixture of some college & bootcamps
Do different education types have different compensation distributions?
Could compensation distribution differences be explained by experience, rather than education type?
Hover over "bootcamp" in the legend, and observe that the points are all clustered between 0-5 years of experience. While bootcamp graduates have a lower compensation when compared to other education types in general, they are within the same pay range as their peers at that experience level (0-5 years).
Similarly, the Graduate Degree responses have the highest average salary, but observe that they are mostly in the 5-15 years experience range. While there is some overlap with the other categories, I don't know if there is enough data to definitely say education type is a strong variable in total compensation.
What type of degrees do respondents have?
While Computer Science Bachelors and Masters degrees remain the largest type of degree held, as a whole, they are a minority to non-computer-science degree holders.
As a category, "other" is larger than CS, MIS, or Business degrees. The degree types vary widely, but include:
- Entertainment Business
- Masters Social Work
- Biology (2)
- Biomedical Engineering
- Exercise Science
- Chinese Literature
- Communications (2)
- Digital Media (3)
- Electrical Engineering
- English (3)
- Graphic Design (2)
- Geology (2)
- History (2)
- International Studies
- Linguistics (4)
- Mathematics (2)
- Mechanical Engineering (2)
- Political science
- Physics (2)
- Psychology (4)
- Supply Chain Management
Where do respondents live?
Where do respondents work?
To hopefully avoid being a "geographic profile map, which is basically just a population map", here are the results in tabular form
What about remote workers?
4 respondents reported living remotely, meaning they live outside of Utah, but participate in UtahJS.
10 respondents reported working remotely, meaning they work outside of Utah.
What are common benefits?
Respondents were asked about other benefits they thought were particularly good. There were many responses, which were quite varied. Most common were:
- Fully-paid health insurance (aka. employer paid premiums) (46)
- Unlimited PTO* (48)
- Matching 401k (20)
- Good* paid parental leave (18)
- Onsite gym
- Onsite health clinic
- Flexible schedule
- Work from home
- Paid or Catered Lunches
- Education reimbursement
- Gym membership reimbursement
* Jury is still out if this is a benefit for the employer or the employee.
* What people considered "good" varied. Most responses were between 4-6 weeks, but several were in the 3-4 month range of paid parental leave.
How many people received a raise this year?
138 respondents reported receiving a raise this year, either at their current employer or by moving to a new employer (raise via moving jobs).
How does receiving a raise at your current employer compare with moving to a new employer?
This shows the distribution of the responses in terms of dollar amounts. However, a comparison of raises as a percentage of their salary may be more useful:
Previous Years Results
This survey was conducted in 2017 and 2018 with similar questions. Those datasets and analysis can be found at:
- 2018: https://mdjasper.github.io/utah-js-pay-data-2018/
- 2017: https://mdjasper.github.io/utah-js-pay-data/
Caveats & Biases
Some of my thoughts about this survey, and on ways this survey could be improved
There were 208 responses, out of 3000 people in the #general channel (in Slack). This is about 7% of members. For this number of respondents and population size, we could say that we have a 95% confidence level, with a margin of error of about 6%. Not the best, but not the worst either.
Members of UtahJS self-selected into responding to the survey, and were not randomly sampled. It's not possible to know if there is a segment of UtahJS that was less or more likely to repond. I believe that there is good representation across years of experience, education types, and job titles. However, the representation of women in the survey decreased from last year. I don't know if a smaller percentage of the overall women participated in the survey this year, or perhaps the number of women who participate in UtahJS declined. In either case, the gender representation was not very diverse.
One particular bias that I would be concerned about in survey responses is survivorship bias, especially as it relates to education data. For example, if an individual who attended a bootcamp, and for a short time participated in UtahJS, but for whatever reason stopped participating (couldn't find a job, moved locations, etc), this persons experience would not be captured in the survey results. Or to put another way, the only data we have is for individuals who are successful enough in their careers to continue.
Because of this bias, we need to be careful about how we understand the survey results. It would not be correct to say that "people who are self-taught make $$ salary," rather "people who are in UtahJS who were self-taught make $$ salary," because those no longer participating aren't here to report their data.
Checkout one of my favorite-to-recommend articles for more info on survivorship bias: https://youarenotsosmart.com/2013/05/23/survivorship-bias/