Why do architecture salaries seem, anecdotally, to be so much lower than statistics indicate?


I guess my question comes down to this: the BLS data on architecture salaries comes out to about $74,00/yr as the MEDIAN salary.  A lot (but not all) architects I talk to/interact with online seem to think its a lot less.  Personally, I see no reason to doubt the BLS figures (collecting and analyzing employment data is basically their only job), but I wonder what could be causing the disparity?  Thoughts?

Jan 28, 13 9:31 pm

How about giving us a link to the BLS data you're referencing?

Jan 28, 13 9:58 pm

I'm pretty sure it has to do with the legal definition of an architect.  The BLS probably only factors in licensed architects and not people working towards their license.  That means they are only sampling "architects" with 11-40+ years of experience since, on average, it takes 11 years to get your license.

Jan 28, 13 10:52 pm

A (kind of) better way of gauging salaries is using or because you can see average salaries by location, experience, job title, company, etc.

Jan 28, 13 10:54 pm

i doubt and are better able to create a reliable data set than the united states government.

Jan 29, 13 7:42 am

...and this is why we need the Archinect salary poll!

Jan 29, 13 7:44 am

The numbers are skewed upward because of confusion and inclusion of software and IT architects.

Jan 29, 13 9:20 am

Where is the damn salary poll on Archinect! I miss it! 

Jan 29, 13 9:41 am
won and done williams

i doubt and are better able to create a reliable data set than the united states government.

As someone who works with data almost everyday, I would take and as more accurate for most pay-related employment data over public sector sources. Massive government data sources are rife with inaccuracies for this type of data.


Jan 29, 13 10:22 am

i doubt and are better able to create a reliable data set than the united states government.

That is really, genuinely funny. You got a huge laugh out of me for that one.

Jan 29, 13 12:13 pm

i've been trying to figure out how a website with user-submitted information and automated analysis is more accurate than a targeted survey.  do you think the people surveyed by the government are lying?  do you think the people who submit profiles to are really an accurate population sample?

or do you think government is lying about their dataset because government is always evil?  i think the methodology for the BLS is similar to that of the AIA salary report, which i think is probably as accurate as you're going to find.

won and done, why would you prefer user-submitted data from the internets?  is anonymity preferable to targeting a source?  or is it just more fun to work with larger data sets?  do you actually think they're more accurate, regardless of the source?  i know there are inaccuracies in the government's statistics, but there are also inaccuracies in payscale's and glassdoor's statistics.  i would add that the government economists are more likely to have a human being differentiate between 'archtiect' and 'java architect'

i will include links to the government's and payscale's method so we aren't making dumb assumptions;

Jan 29, 13 1:51 pm

aia's methodology if you've been having lots of fun looking at how statistics work and don't want it to end:

Jan 29, 13 2:02 pm

Does anyone have access to/ willing to share the full AIA compensation report? 

Jan 29, 13 3:02 pm

Curtkram, I appreciate the links to the methodology, though to be honest, I have no real way of knowing whether one is any better than the other.  I'm inclined to trust BLS data, because as a general economic indicator it is incredibly accurate, but of course that might change on a case by case basis.

Basically, there are two "camps" of surveys: the self-reported, online surveys that state an average income of about $56,000 (yikes!), or the professionally compiled and statistically adjusted surveys which state the average income to be about $73,000 (nice!).  While I'm personally inclined to assume that these surveys are much more accurate since they don't suffer from self-selection or self-reporting biases, I also wonder if they are only considering registered architects, whereas the online ones account for all people in the profession.  Does anyone have any insight?

Finally, the AIA compensation report seems to divide employees into the categories of Intern I, II, and III, as well as Architect I, II, and III.  Does anyone know what the definitions for these categories are?

Jan 29, 13 3:22 pm

Fun little resource, though obviously years experience in addition to job title would be much more helpful:

Jan 29, 13 4:42 pm


here ya go - position descriptions:

While dated 2005, I think these remain more-or-less current.

Jan 29, 13 4:43 pm

aphorismal, that's a good question about intern v. only registered architects.  i don't know the answer.

here's a link from cooper union with part of the 2008.  i replaced '.' with 'dot' because i think it might be a bit shady that they have this on the internets.


i didn't put that on the internets, i'm just disseminating what other people put up.  if the mods want to edit that and leave this one behind that's fine too:

Jan 29, 13 4:52 pm

Thanks guys, that's very helpful!  I'm going to go out on a limb and guess that "first year of internship" = completed less than 1/3rd of their IDP hours, not "been an intern for 1 year."  Still, all things considered, that's not too bad.

Jan 29, 13 5:11 pm
won and done williams

I prefer sources like for the reasons they cite in your link above, specifically user-generated data brings the data closer to the primary source, i.e. the employee, than the BLS which has to aggregate from multiple sources. There is a nice public/private sector analogy here: in order to find out what the public thinks, the public sector creates an elaborate bureaucratic structure to get the answer; the private sector simply asks the question.

Jan 29, 13 5:25 pm

won, that makes sense on a certain level.  But the reason BLS aggregates from multiples sources isn't just to support an elaborate bureaucracy; it's to create a statistically accurate reflection of reality , or at least get as close as possible.

Also, the private sector "simply asking" the public what the unemployment rate is would produce nearly useless results...

Jan 29, 13 5:54 pm


You may find the link below -- which is reachable from the BLS web-link provided earlier today (at 7:42 am) by curtkram -- to be useful in your quest for clarification. In particular, it is more descriptive of the job requirements included, plus it allows you to refine the BLS data by State and provides more precise quartile data

Unfortunately, neither the above link nor the BLS link provided by curtkram is totally explicit about whether the data is only for licensed architects. However, the BLS page seems to imply that to be the case -- if you follow some of the links provided there you come across this phrase: "All states and the District of Columbia require architects to be licensed." (

I tend to believe the BLS data would be highly reliable data because of governmental reporting requirements. However, unlike the AIA Compensation Survey - which has the attribute of collecting data for specific job levels - the BLS data is aggregated for the whole pool of workers the BLS considers "architects".

Hope this helps.

Jan 29, 13 9:18 pm
won and done williams

the reason BLS aggregates from multiples sources isn't just to support an elaborate bureaucracy; it's to create a statistically accurate reflection of reality , or at least get as close as possible.

I didn't say the BLS does what it does to support an elaborate bureaucracy; that's the symptom and not the cause. My main point was that the further you remove data from the primary source the higher the rate of inaccuracies in the data despite the purported effort to make it statistically more accurate. The archinect salary poll was actually highly accurate despite the sample being relatively small because it was a primary source. It was great data to work with. I will take data from 10 interns sitting around over a few beers over all your government data sources any day of the week to get a real sense of a local market.

Jan 29, 13 10:24 pm

Actaully, quizzical, I don't necessarily agree with you.  The main reason is because the BLS figures match the AIA ones very very closely, and since one takes interns into account, I would assume that the other does too.

But the main reason is their classification system; EVERY employee in the country is somehow categorized by the BLS (seriously; they have thousands of classifications:, and most of those categories are defined by the responsibilities and tasks of the employee in question.  The only two which could even REMOTELY include architectural interns are:

17-3011 Architectural and Civil Drafters (Mean Salary: $50,000)

17-1011 Architects, Except Landscape and Naval (Mean Salary: $79,000)


I'm going to assume that, unless their employers count them as drafters, architectural practitioners (licensed or otherwise) are counted in the second category.

Jan 29, 13 10:29 pm

Won, I'd love a link to the archinect salary poll if you've got it!

And in principle I agree with you about over-processing data, but on the other hand, all the archinect salary poll accurately reflected was the salaries of architects who frequent archinect.  I doubt that sample reflects the whole profession; for one, this site has a somewhat avant-garde and sculptural/installation-work lean to it, which I would guess pays less.  For another, not trying to hurt anyone's feelings or name names, but the less work you have to do/the less employed you are, the more likely you are to spend time bullshitting around these parts (myself included ;P).  I guess what I'm trying to say is that sometimes the easy data is also the most biased; all 10 of those interns drinking beers and telling you their salaries might give you a good view on the local market, but they might also all be in the same network of small, low-paying firms.  The interns drinking at the next bar over could be making significantly more, but you'll never hear about it.

Jan 29, 13 10:39 pm
won and done williams

aphorismal, your post at Jan 29, 13 10:29 pm points to the exact reason why the BLS methodology is flawed. "17-3011 Architectural and Civil Drafters"? How many of those do you know? In my world, we call them interns. I disagree with your assessment of localized small-sample data - the difference in pay between your 25-year old 60-hour/week Rhino-scripting whore and your 25-year old facebook-surfing, archinect-posting slacker is negligible within a localized market; a sample of about 10 will give you a highly accurate portrait of the local salary range. Anyway, sorry to pull this in such a nerded-out data discussion.

I think the salary poll was pulled when the website went to 2.0 (or is it 3.0?). I haven't seen it recently anyway. Even if it were available, it's probably a little dated now; most of the posts were from 2005-2007 as I remember.

Jan 29, 13 10:58 pm

The only positive about my salary is that I have a job. And the only positive thing about my last raise is that they aren't ready to let me go.

Jan 30, 13 3:05 am

Block this user

Are you sure you want to block this user and hide all related comments throughout the site?

  • ×Search in: