There is an Employer address, and then there are two separate fields for Working Location. So part of the problem is are we looking at where the company is registered, or where the employee is physically working?
A second issue is in the 'LCA_CASE_SOC_NAME' - the job title / category. I found 27 categories which are reasonably related to computer software field, including things like 'Graphics Design', 'Computer and Information Systems Managers', 'Network Systems and Data Communications Analyst', etc. If you filter on just the specific 'Computer Software Engineers, Applications' category you will see much less.
The 2012 file shows decisions made 10/1/2011 thru 9/28/2012. If I look at data for 'Employer State' of OR, and filter on 27 categories which seem computer/software related, then I see 321 records. If I look at data for 'WorkLoc1_State' for OR with the same categories, I can see 1557 records.
You can download it yourself here [1] although it's 70MB and takes several minutes to pull from their servers.
[1] - http://www.foreignlaborcert.doleta.gov/docs/py2012_q4/LCA_FY...
I looked at the website briefly, but wasn't sure how he was able to aggregate all of the data.
- Prevailing wage
- My actual base salary (higher than above)
No sure which of the above they're using for this infographic (I know the actual base salary has to be >= prevailing wage, but is it public?), but when you include stock grants / bonuses / 401k matching, my total compensation is well above both of those numbers.
http://www.flcdatacenter.com/download.aspx
older data here:
Drive down homegrown market prices by increasing supply.
https://my.infocaptor.com/dash/mt.php?pa=h1b_metrics_5224efd...
"In 2012 the Oregon software industry gave jobs to 38 foreign software engineers."
Ok so here I am assuming you mean 38 new H1B engineers were hired in 2012. There is no way that could be true, I nearly hired and processed that many H1Bs personally in 2012. I was working with several others with similar or higher numbers. And that's just one company in Oregon.
Not trying to nitpick with anecdotal data, but it just seems wrong to me.