A recent study showed that software developer employment peaked in 2019 and has been declining since. I question some of the methodology and conclusions.
Shameless plug
Before I continue, I’d like to discuss a new project I am working on. It is called Ignorance and Confidence Podcast. I partnered with Mark Porter, a good friend who happens to be a tech recruiter. The concept is simple: Mark is a recruiter, and I am a seasoned software engineer. We discuss careers, work-life balance, market conditions, expectations, frustrations, and personal and professional growth. All opinions expressed in this podcast, like in my newsletter, are those of the authors and contributors and do not represent anyone else.
Podcasting is new to me. I always thought of myself as more of a writer than a speaker, though I have done several public speaking engagements. I fully expect to make mistakes, misspeak, say too many “Uhms,” and have racing thoughts. I am okay with that 🙂
You can follow the “Ignorance and Confidence” podcast on Substack, as well as find it on Apple Podcastsand Spotify.
Are developer jobs indeed in decline?
It just so happens that we touch on this topic in the podcast pilot episode. Here are the two broader points to consider.
First, what does it mean to be a developer?
The study identified “developers” as follows:
A set of employees was identified by querying a set of keywords present in known software developer job titles (such as software engineer, C++ developer, stack developer) and querying O*NET occupation codes for software developers (15-1252.00, 15-1253.00, 15-1254.00 and 15-1221.00).
I looked up the occupation codes. Here is what they are:
- 15-1252.00 : Software Developers – Includes software engineers and architects
- 15-1253.00 : Software Quality Assurance Analysts and Testers -Includes testers, business/IT analysts, and test engineers
- 15-1254.00 : Web Developers – Includes roles like “webmaster”
- 15-1221.00 : Computer and Information Research Scientists – An odd niche category to include, but includes computer scientists, research scientists, and scientific programmers
My take on this is that the nature of being in any of these categories has evolved in the past decade. Many occupations now require the ability to write code. There are new fields like data scientists, computational physics, computational finance, and computational biology that aren’t easily classifiable as “software development” roles. Among the examples I just mentioned, only “data scientists” show up in O*NET OnLine database.
Is someone writing code to simulate biological processes a developer or biologist?
The rise of new specific fields means the decline of broader generalizations like “software developer.”
Could it simply be a market correction?
If you look at the chart citing the decline in software developer employment, there is a peak in 2019, followed by a crash in 2020 (i.e., the pandemic and market uncertainty due to the Jan 6th, 2021 events) and over-hiring in late 2021 and 2022. It does not look as drastic if you flatten that curve to compensate for the over-hiring during the pandemic.
Combined with the changing nature of what it means to be a developer, as I described above, I don’t see how a conclusion can be drawn that software development employment is on the decline.
Final thoughts
Ed Yourdon talked about the decline and fall of American programs in the 1990s, to write a second book about their rise and resurrection. Our field goes through periods of expansion and contraction. Each cycle generates new ideas, new technologies, and new classes of jobs. Whether you call yourself a “developer,” “computational physicist,” “data scientist,” or “business intelligence analyst,” your future is bright. I wouldn’t worry.
The Dulin Report 