2017 Design Census: Participation
An introductory look at AIGA and Google’s newly released data-set surveying designers in 2017.
For a six-week period during November and December 2017, AIGA, the professional association for design, and Google partnered to survey the US Design Industry in their second annual Design Census. Just over 13,000 people participated, and the information they gathered is now freely available for all to explore at https://designcensus.org/
Before diving headfirst into the finer details of the data-set it’s important to orient ourselves with context. A high-level overview of who participated allows us to be aware of both strengths and shortcomings present in this year’s results. Doing so can help shield us from our biases and our temptation to cherry-pick from the data.
Total Participants (before data cleaning): 13,158
Total Participants (after data cleaning): 10,257
A complete data-cleaning methodology can be found at the end of this piece.
2017 Design Census, Total Participants, US Regions
Regional participation numbers reveal encouraging facts:
- Participation is higher across all Regions compared to 2016. Great!
- The ‘West Pacific’ region represents 20.5% of the data set, with ‘Northeast Middle Atlantic’ close behind with 18.1%. Conclusions drawn from the data are statistically weighted in favor of these regions.
- Both the ‘Midwest East North Central’ and ‘South South Atlantic’ received 1500+ participants. This helps spread some (but not all) of the data’s weight. To illustrate: In 2016 there was a 10.3% difference in participation between the first and fourth most subscribed regions. In 2017, that same difference is only 5.5%.
2017 Design Census, Total Participants, US States
State-level participation numbers reveal discouraging facts:
- Totals in CA and NY (26% combined) eclipse participation in other states by an outstanding margin. Great if you’re in CA or NY. Less so if you’re not.
- A total of 24 states each collected fewer than 100 participants (and 18 of those states had fewer than 50 participants).
Average Salary Breakdown
Figures 2A and 2B:
2017 Design Census, Average Salaries Heatmaps, US Regions & States
Being mindful of the fact that year-two data won’t yet reveal design industry gospel truths, the 2018 Design Census Salary Averages suggest:
- Of the cleaned data set, the national average annual salary for a Designer is $69,466. It’s critical to see this number as a starting point. Design covers a wide array of skills and disciplines at varying levels of compensation.
- CA, WA, and NY States boast the highest average annual salaries, ranging between $81k — $95k. We know that CA and NY represent a substantial portion of the data set, which affirms that these numbers aren’t unbelievable.
- There is no correlation between average annual salary and participation levels. This is a good thing. If participation was perfectly correlated across all census categories it would prove the data unreliable.
Breakdown of Non-White Participants
It does a disservice to the topic of racial diversity to be so binary, but the term ‘non-white’ is suitably descriptive of what data is being visualized in Figure 3.
Figures 3A and 3B:
2017 Design Census, Total Non-White Participants, US Regions & States
A high-level view of Design Census data on race and ethnicity shows:
- 27% of the data set are non-white. In simpler terms: for every 1 non-white participant there are 3.8 white. This is slightly higher than national population estimates shared by the US Census.
- At a state level, 18 states have 27% or more non-white designers in their communities, but 33 states have fewer than 27%. This is partially correlated with census participation levels.
- Of States with 100+ participants, FL has a noticeably high 44% of its total participants identifying as non-white Designers. This doesn’t necessarily mean anything, but it’s an example of possible stories in the data that warrant further investigation.
Breakdown of LGBTQ Participants
Figures 4A and 4B:
2017 Design Census, Total LGBTQ Participants, US Regions & States
Examining segmented data of people who identified as members of the LGBTQ community suggests:
- 11.7% of the data set identify as members of the LGBTQ community. Which translates: For every 1 LGBTQ participant there are 8.6 non-LGBTQ participants. National population data on sexuality is not yet concrete, which makes it difficult to compare our number to widely reputable sources.
- 1197 LGBTQ participants makes for a fairly small sample size. As social discussions around representation and inclusion continue it will be important to identify the ways this data should be developed and grown.
It’s important to note when studying today’s Race and Ethnicity and LGBTQ data, that the numbers cannot yet be considered to fully represent these communities, and should only be viewed as a snapshot. Diversity in Design is a leading area of focus for AIGA and other design associations active today, so we can assume to see these participation totals grow in future years. There is work to be done here.
Conclusion: Just the Beginning
Speaking personally, as a designer and AIGA chapter President, it’s easy to notice the changing landscape that the design industry and its communities experience every day. This growing Design Census data is a compelling catalyst for important dialogues and questions Designers should be having.
This piece only scratches the surface, but 2017's data presents encouragement and challenge for both the surveyor and the surveyed. Although it’s exciting and satisfying to see the Design Census grow in maturity and size compared to its 2016 sibling, there is a mounting need for the survey to advance in its ability to consistently reach Design communities across the US.
Thankfully, AIGA have made a long-term commitment to being a leading source of design industry data, and the Census is one of two major efforts in 2018. The second is a study called “The Designer of 2025” — a view of meta trends affecting designers today — and will be available later this spring.
In a similar piece compiled for last year’s census, I concluded “even with a sample size of 100,000 it would be difficult to make definitive conclusions from this information until there’s 3–5 years of data to draw accurate trends from”. This remains true, but not as a mark that devalues the importance and significance of the Design Census initiative. As data for years three and four are gathered it will be imperative to focus on reaching areas currently under-represented in the data.
In the meantime, we need to learn clearly where the under-represented areas of the data are. Download it here and start exploring: https://designcensus.org/
Data Cleaning Methodology
- Creating location heat-maps is dependent on participants answering the optional census question to provide their zip-code. Of the 13,158 who participated 2,382 (18%) chose not to share their location.
- Of the 306 entries missing annual salary information, only 1 claimed to be currently employed. The remaining 305 data points were given an annual salary of $0.
- To help remove outliers, all participants who fell into the 1st ($0) and 99th ($216,000 and above) percentiles of annual income were cleaned from the data — a total of 518 people.
- After cleaning, the final data-set contained 10,257 entries.