Last time we were looking at salaries for people with the job I have, and comparing them to expectations one might reasonably infer from reading the University's HR website.
Today I'm going to look at the job (notionally) I want: University Professor.
There are 5 publicly funded universities in Washington State. Here is how the 2017 salaries for people whose job contains the word "Professor" compare:
There's a lot to unpack in this graph, even with just 5 curves...
- UW is the clear "winner", with a median salary (vertical bar) far above any of its "peers"
- UW has a TON of faculty (and/or their faculty have more line-items in the budget - i.e. a single Prof having multiple entries)
- The 4 other schools are quite closely clustered around $80k-ish
- The 3 schools located east of the Cascade mountains (i.e Eastern, Central, and Washington State) all have very similar primary colors listed on their websites. Coincidentally, these schools all reside in "red" districts. This color proximity drove several design choices for the viz.
But comparing average salaries does not tell the whole story...
These schools are located in vastly different regions of our state: a major city, a rural farming community, a ground transportation hub, an international boarder... and life in each of these cities/towns is equally unique. So for your consideration, here is the same data as above, but normalized by the median home price for each city:
Wow
- Given the (ridiculous) cost of living in Seattle, it is no wonder so many faculty now have to live outside the city. For example, I live ~12 miles north of Seattle where home prices are ~27% cheaper.
- Bellingham has gotten expensive!
- Faculty in eastern Washington are doing substantially better than their western counterparts... probably much closer to what being a Professor in most cities used to be like (i.e. buying a reasonable house near the University on a faculty salary)
This is an age-old debate when considering the job market for faculty... should one chase the cosmopolitan lifestyle of a big city, or be in the upper echelon of a small town? Clearly you shouldn't just look at top-line salary when considering which university to work at. While there's no "right" answer, I for one find the high earning power of rural faculty quite promising. Small towns can be wonderful places to live, and competitive salaries can bring top talent to these schools.
One more thing...
Washington state has 30 public colleges (mostly community colleges). Here is the data for jobs listed as "Professor" or "Faculty", with medians shown as heavy circles....
Most of these curves are very skewed towards low salaries, endemic of the state of college faculty hiring and the reliance on part-time labor...
Of course, all the Python code to do this analysis (mostly just bread/butter Pandas) and make these figures (matplotlib) is available on my GitHub profile.
Does the data source allow for determining the professor's field? This analysis might look very different if you are comparing the salaries to the median salary for that field outside of university. Even just by college (i.e. Engineering vs. Business vs. Liberal Arts vs....) might be enough to be informative.
ReplyDeleteThe variance between those salaries is extreme, and I've never seen it accounted for in any study, and I've seen a lot. Liberal arts professors are paid WAY less than marketing and law professors. Half as much in lots of cases. I had a tenured wife that worked about 20 hours a week, mostly online "work" as a marketing professor at a satellite school for the U of Indiana and she made around $110k/yr. for 10 months, not including some very generous vacations, policies and rather undeserved stature. It's a cush job, for sure.
DeleteWhat is the Y-axis supposed to represent?
ReplyDeleteThey are stacked vertically skewed stacked histograms, (which many folks in the statistics business wrinkle their noses at for providing a source of confusion). That is, there's some unstated sense of binned data wherein one may say this range of salary values represents some percentage of the whole sampled population. The highest point of each curve is thereby the mode salary, etc.
Delete