Yesterday I looked through the CareerBuilder.com national salary data that South Dakota top-tier ad firm Lawrence & Schiller used for an ad campaign for the Build Dakota Scholarship and found that the data show South Dakota paying lower wages than the national average for almost every job you can think of.
To create my table of average salaries for South Dakota and the U.S., I entered 41 job titles into Career Builder’s salary calculator yesterday, Friday, June 17, 2015, between 06:00 and 07:30 CDT. I returned to the same webpage and entered exactly the same job titles this morning, Saturday, April 18, 2015, between 09:45 and 11:15 CDT. You can see and play with my data on this Google spreadsheet.
Within the 28-hour span between my two sets of queries, all 41 national averages changed. 22 went down; 19 went up. Disregarding the plus or minus sign (looking at absolute values), those 41 national average salaries changed 1.66%.
Fewer of the South Dakota average salaries changed, but some of the changes were whoppers. 15 stayed the same, while 14 went up and 12 went down. The average absolute change was 3.84%. From Friday to Saturday, Career Builder’s reported pay for South Dakota bankers dropped $4,083. South Dakota programmers’ pay went up $9,536. Pharmacists went up $11,042. Nurses went up $24,651, meaning that while South Dakota nurses were paid 27% less than the national average Friday, they are getting paid 12% more than the national average today.
There could be some logical explanation for these changes… but all Career Builder tells us about their average salary calculation methodology is that “Data is sourced from Careerbuilder’s robust salary information housed in our Compensation Portal.”
That sounds like corporate code for, “Our numbers are baloney, and phooey on you for asking.”
“Robust salary information” for professions including thousands if not millions of workers would not fluctuate that much from day to day. Such data would be sturdy enough to withstand a handful erroneous reports or job listings showing anomalously high or low salary offers.
As I noted yesterday, Lawrence & Schiller could have built its ads for the state’s new vo-tech scholarship program around more rigorous economic data, like Bureau of Labor Statistics wage data, which does not change from day to day. Instead, Lawrence & Schiller chose a commercial website using unknown methodology that produces different data for the same queries on Friday and Saturday. These sizable daily fluctuations indicate Career Builder’s data is neither robust nor reliable.
We can thus more firmly conclude that Lawrence & Schiller, our state’s preferred advertising vendor, used our tax dollars to create a marketing campaign based on crap data.