Add jobs data by commodity

[coreycaitlin]

We don't have jobs data by commodity at the county level, but we do have it at the state level. We could consider showing state-level jobs data by commodity in addition to general extractives jobs by county.

[usEITI]

I think it would get us a little closer to Betsy's vision. It would be an improvement.

…

On Wed, Nov 23, 2016 at 6:14 PM, Corey Mahoney ***@***.***> wrote: We don't have jobs data by commodity at the county level, but we *do* have it at the state level. We could consider showing state-level jobs data by commodity in addition to general extractives jobs by county. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#2156>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ANnOZHsi_G1qj4NS2Pl_S_ei2d225hrZks5rBMi9gaJpZM4K7IpJ> .

-- *Judy Wilson* *Program Manager USEITI Secretariat * *Office of Natural Resources Revenue* *judith.wilson@onrr.gov <judith.wilson@onrr.gov>* *202-208-4410*

[coreycaitlin]

Within Wage and salary jobs, we'd like to have two sections instead of one. The first (what's there now) could be called "Jobs by county;" the second could be called "Jobs by industry."

The "Jobs by industry" data isn't county-level, just state-level, so it would be much like this:

[coreycaitlin]

For the nav, we'll probably need to pull wage and salary out of the title of the section, and let the in-section content cover that.

Thus, the right-nav for this section would list:

• GDP
• Jobs by county
• Jobs by industry
• Self-employment
• Exports

[coreycaitlin]

see #2168 for prior discussion of which industries we're including.

[coreycaitlin]

Upon further investigation, it looks like we might have county-level industry data after all! More next week...

[shawnbot]

Okay, so I'm working on updating our data at both the state and county level to include the industries listed in #2168. This called for a total rewrite (and simplification) of what we were doing before. Here are my tasks, marked up as subtasks in #2303:

• Update data scripts to extract, transform, and load data by commodity
• Update templates to render state-level data by commodity
• Update YAML files to output commodity data by state and county
• Validate/test

[shawnbot]

Well, I'm working my way through this, and I think I've hit a blocker: the BLS data for some standout counties is wonky.

Using Eureka County, Nevada as an example: the latest BLS data for 2007 (search for "eureka") lists 3,092 jobs in Gold, out of 5,025 total jobs:

However, the 2008 data (search for "eureka") lists 4,452 jobs, but none in Gold or "Metal ore mining" (note: we don't output rows with zero jobs):

Our old 2007 data — which represents extractives totals — matches:

And, surprisingly, our 2008 data has zero rows for that county. So maybe it's not a blocker in the sense that the data hasn't changed since we last pulled it, but that the county-level data is unreliable?

[shawnbot]

Okay, so it looks like the state data is good, and we're going to roll with that. County extractives jobs are definitely wonky, which you can see in Nevada when you switch between 2007 and 2008: