Comment by srl | Hacker News Reader

Comment by srl | Hacker News Reader

Parent Full thread

This is a huge effect. Implausibly huge? I'm reminded of Gelman et al's piranhas: http://www.stat.columbia.edu/~gelman/research/unpublished/pi.... The "folk theorem" version is that you can't have very many independent variables that all have large effects on salary. Saying that spaces vs tabs has a 10% effect is tantamount to saying that its one of the ten (or fewer) most important things about a developer.

So there must be a confounder. What could it be?

- Do the top schools all instruct students to always use spaces? This should have been partly taken care of when the author controlled for level of education though, right?

- Choice of editor? This seems like a weird one, but editor defaults are definitely linked to tabs vs spaces, and the survey didn't even collect data on this.

- Starting language?

- Total number of languages used? Spaces are tolerable in more languages than tabs are. Of course, I'd expect "both" to beat "spaces" if this was the case.

None of these strikes me as particularly plausible, except maybe the last one. This result is kind of bothersome, because it's big, obviously not a statistical fluke, and hard to understand.

EDIT: a past thread (https://news.ycombinator.com/item?id=20719010) suggested that some higher paying companies mandate spaces. This also seems plausible.

Only thing I can say is that I started my career with tabs as it made more sense on the surface, but then over time migrated to spaces which did not have a lot of issues that I was encountering with tabs, while interoperating with other projects,etc. Spaces are much more frictionless than tabs, so maybe the attitude of trying to reduce friction would be such a variable ?

> I started my career with tabs as it made more sense on the surface, but then over time migrated to spaces which did not have a lot of issues

I think this is a common phenomenon, not just for individuals but for entire organizations. Over my career I have entered various large existing software projects. I've only seen two classes: (1) "we use spaces only", and (2) "we used to allow tabs but new code must use spaces only". Whatever the perceived theoretical benefits of tabs may be, in practice they seem to cause headaches for teams.

> I started my career with tabs ... but then over time migrated to spaces

Perhaps this anecdote resonates with many developers? If so, then perhaps tabs vs spaces is correlated with years of experience, which is correlated with salary.

My qualified guess is that PHP developers overall tend to use tabs more, and PHP developers are generally paid less on average.

If you read the article closely, you'll see that they made some effort to control for programming language, and in particular they note that this effect occurs within each subpopulation. So, among PHP developers only, it's still the case that spacers make substantially more than tabbers.

Very common in the Bay Area that spaces are mandated. Personally, I don’t like it but it’s what I see here. If you included what I normally do and at the comp I get, just one Bay Area participator could fuck up the stats real fast.

It was already discussed in the past, one major source of the difference was that in Europe tabs are more common and it is also more common to talk about monthly salary,

Reply on news.ycombinator.com