Explore > Jobs: numbers are off #1092
Comments
shawnbot
added
the
data-mgmt
|
Boooooooo! |
|
I think Isabelle is out for a little bit. This is the latest info I have: Me: What do you think about the wage/salary numbers? And I'm thinking the self-employed problem is just a order-of-magitude problem. Isabelle: |
|
Hi I'm here! I'm looking! |
|
Ah 1.65 million is self-employed + wage and salary employment |
|
You likely have SA25N which is good. Now you just need to subtract SA27N from it to get a "self-employed" total. I'm not looking at Shawn's data, just going on messages for now. I'll try and revisit the data and give you some sample numbers. |
|
For the U.S., I'd expect 797,000 (in the exec sum, it's 799,000, but this is from a data release before they updated it in November, so don't worry about that discrepancy); for Alabama, I'd expect 5,672. For Alaska, I'd expect 4,271. This is just for self-employment, not total employment or wage and salary employment. |
|
And yes, looked at site and numbers for "self-employment" seem to be for all employment. Best, Isabelle |
|
@Isabelle1512 do the numbers look accurate for wage and salary then? |
|
They look ball park but slightly off. If I download data from BLS for NAICS 21 (you may or may not be also adding renewables, but those are just a few thousand) - I would expect 289,610 for texas (not including renewables). I gave Shawn a bunch of codes, but that was mostly because we said we'd break out jobs by commodity. If we aren't breaking out jobs by commodity, then to get totals you just need NAICS 21 plus the renewable codes which all start with 22. Off to a meeting, will log on later tonight. |
|
Ah ok, I forgot to include the NAICS 22 jobs in the BEA data. So that's first on my list. |
|
Okay so, based on the Texas figure at least, I think wage and salary numbers are right now. @Isabelle1512 can you check out this preview link and/or send me the spreadsheet that you're referencing so I can check them on my end? |
|
@Isabelle1512 I think I've got the self-employment figures fixed now too. The US total shows 797,000 and Alabama shows 5,672. |
|
I had to go back and look at #923 to be sure, but the deal is that the BEA data isn't organized by NAICS codes, and their "line codes" don't appear to align 1:1 with NAICS. We're getting 200 ("Mining") then getting 10 ("Total employment") so we can derive percentages for #955. These are all of the codes, according to BEA's API. @Isabelle1512, do any of these stick out to you as relevant, or do the numbers in the preview links above look right now?
|
|
FWIW, the self-employment issue was a double-counting problem. The data contains figures at the national, state, and county levels, and I was adding them all up. |
|
For reference, these are the tables in the executive summary for wage and salary: ...and these are the tables for self-employed: |
|
Okay, I think we're good here when #1103 merges. On that branch we list 813,040 wage and salary jobs, which matches up pretty closely with the report's 808,000 (page 81):
The self-employment figure in the report is confusing. It cites the BEA data listing 16,000 self-employed persons in the table:
But then it goes on to say:
On my branch we list the nationwide total for self-employed persons as 797,000. I'm honestly not sure how to reconcile the slight differences, but my gut tells me that they're close enough. |
|
@shawnbot, Isabelle did say something about the numbers getting updated every month. That might be the discrepancy. |
|
This should be live on the site now. @Isabelle1512 if the numbers look good to you we will close this issue. |
|
@shawnbot and @meiqimichelle : The self-employed 16 K in the report is what they call "self-employed." If you remember from the BEA data, their categories are self-employed (16K), "wage and salary" (800K) and "employment" (1600K). We backed in to a new "self-employed" total (800K) that includes sole proprietors and active partners. The agreement reached in the Executive Summary was to discuss all three: self employed, wage and salary, and self-employed inclusive of sole proprietors and active partners. The agreement we reached in the online report was to do both wage and salary and self-employed inclusive of sole proprietors and active partners. It's all really confusing, and the product of less than perfect data (the BEA had trouble explaining some of their methods to our stakeholders) and a grand bargain on how to discuss employment in extractives. Ideas for how to make it clearer in 2016 welcome! |
|
"they call" meaning the BEA |
|
The self employed inclusive of sole proprietors and partnerships in the Exec Sum uses the previous data release, so don't worry about small differences. I hadn't realized how frequently they updated historical data. We'll be more strategic and clearer about how to cite that in the Exec Sum next year - perhaps by rounding more so that their aren't these small differences with the online report which will be able to use the freshest data I'd imagine. |
|
their --> there. No coffee yet. |
|
Not sure if I'm looking at the right version of the site, but here are the values in yellow I'd expect to see. |
|
This was fixed before launch on Tuesday. Done! |
Our numbers aren't matching up. @Isabelle1512 reviewed this page in the past and life was good, but it looks like we introduced a regression and our numbers aren't correct anymore. Isabelle, can you list what numbers we should expect to see on this viz? Thank you!
The text was updated successfully, but these errors were encountered: