Monday, April 22, 2013

2012 ACS Starting Salary Survey: Unemployment, starting salaries down slightly

The 2012 ACS Starting Salary Survey is out. This survey measures unemployment for 2012 graduates; the survey was sent out in October 2012 and collection ended on January 1, 2013. The headline numbers from the article by Susan Morrissey:
  • Unemployment for 2012 ACS new graduates is down 0.7% from 2011, to a still depressing 12.6%. 
  • Median starting salaries are all flat, or down: 
"The largest change was for new Ph.D.s, who saw starting salaries drop to $75,000 in 2012 from $85,000 the previous year. For those earning master’s degrees, the reported median starting salary fell from $55,000 in 2011 to $49,500 in 2012. Starting salaries for bachelor’s degree earners held flat at $40,000."
(The Eka-silicon caveat: The response rate for this year's survey was 16.6%, which is slightly down from last year's 17%. E-S (and myself) would like to remind folks that this introduces a great deal of imprecision to these numbers, to the point that E-S would probably say these numbers should not be commented on.)

A couple of things that I can see right off the bat: I've always challenged the broadly-held notion that "master's chemists are 'more employable.'" For the most part, you can see that, from an unemployment point-of-view, that is not actually true. The trends for unemployment are consistently with lower unemployment correlating to education level, i.e. Ph.D.s have lower unemployment than M.S., which have lower unemployment than B.S. chemists. You can see this same trend holds true for new grad unemployment above. However, it is interesting to note that starting master's chemists seem to have the highest levels of full-time employment above. So, in that sense, the old saw is partially true.

I think it is interesting how many people get pushed into "more school/more training", across the 3 educational levels. If you believe these numbers (a big if), that means that somewhere around 30% or more of each graduating class decides to spend more time in academia. Also, the "not employed/not seeking" percentages are disappointing. It's sad to see 7-10% of each graduating class fall into that category, which does not do good things for one's career.

Finally, an interesting factlet from the article on the salary gap between men and women new grads:
The gender of graduates with less than a year of experience also affected starting salaries. Male bachelor’s degree recipients indicated that their median salary was $43,000, whereas for women, the median salary was $36,500. Similarly for Ph.D.s, the median salary for men was $81,300, and for women it was $74,000. Women who earned master’s degrees, however, reported earning more than their male counterparts—$48,000 compared with $45,000.
There's a good bit of food for thought there.


  1. Don't forget about the salary comparator if you want to test your situation.

    Personally, I'd like to see a bit more interactive data regarding a few data points. Ideally, I'd like to see the salary distribution as a function of years since BS and simultaneously pull out the managerial and non-managerial roles.

    My hypothesis is that I am pretty sure that the managerial salaries are skewing the data a bit, particularly for the older workers. Maybe I'm not taking into account the right combination of categories.

    1. That's a neat salary comparator. I'm in Canada so it's not directly comparable, but it's reassuring (ha) to see my salary fall at about 25th percentile.

  2. The inflation-adjusted salary data just looks like noise to me, with perhaps a barely statistically-significant decline after the crash.

  3. In regards to the gender salary gap, I'd be very interested to compare salaries offered versus starting salaries. While we'd have to allow for variation in achievement on an individual basis, this may give some insight into unconscious bias as well as men and women's ability to successfully negotiate (if at all). As for master's level women, no clue as to why they are bucking the trend, will def be on the lookout reasonable hypotheses (if these numbers are true)