You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Nov 10, 2020. It is now read-only.
shawnbot opened this issue
Sep 8, 2016
· 2 comments
Assignees
Labels
data-mgmtUse this label for adding and/or updating data on the site. Also can be used for data management toop2: HighThis doesn’t prevent the site from being used, but needs to be addressed in the near term.
We should have null values in our self-employment YAML files so that we can correctly identify them as withheld, per BEA's descriptions:
(D) Not shown to avoid disclosure of confidential information, but the estimates for this item
are included in the total.
(L) Less than 10 jobs, but the estimates for this item are included in the total.
Actual behavior
We're parsing BEA's (D) and (L) values as zero, as mentioned in this comment.
This is kind of tricky to debug, because in order to get the self-employment counts we're subtracting the number of wage and salary "mining" (NAICS code 21) jobs from the total number of mining jobs. That happens in our get-bea-data.js script, and I've confirmed that it's producing some of the right numbers by spot-checking the values on the BEA site.
I think that the fix, though, is to treat the self-employment number as withheld if either the total or wage and salary values is listed as (D) or (L). What do you think, @meiqimichelle?
We'll know we're done when...
Delaware should have just 165 (181 - 16) jobs in 2006, something representing the 177 - L value (where L means "less [sic] than 10 jobs") in 2009, and withheld values instead of zero for 2008-2014. Note: Delaware in 2007 is the only instance of the L, so we might be able to get away with just subtracting 10 from 177 and calling it a day.
Main should have withheld values in 2008, 2009, and 2012.
Rhode Island should have withheld values in 2010, 2011, 2013, and 2014.
DC should have values for all years, because it has non-zero values for all employment and all zeroes for wage and salary jobs.
The text was updated successfully, but these errors were encountered:
shawnbot
changed the title
Properly identify withheld values in BEA (self-employment) data
Properly identify withheld values in BEA self-employment data
Sep 12, 2016
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Labels
data-mgmtUse this label for adding and/or updating data on the site. Also can be used for data management toop2: HighThis doesn’t prevent the site from being used, but needs to be addressed in the near term.
Expected behavior
We should have
null
values in our self-employment YAML files so that we can correctly identify them as withheld, per BEA's descriptions:Actual behavior
We're parsing BEA's
(D)
and(L)
values as zero, as mentioned in this comment.This is kind of tricky to debug, because in order to get the self-employment counts we're subtracting the number of wage and salary "mining" (NAICS code 21) jobs from the total number of mining jobs. That happens in our get-bea-data.js script, and I've confirmed that it's producing some of the right numbers by spot-checking the values on the BEA site.
I think that the fix, though, is to treat the self-employment number as withheld if either the total or wage and salary values is listed as
(D)
or(L)
. What do you think, @meiqimichelle?We'll know we're done when...
181 - 16
) jobs in 2006, something representing the177 - L
value (whereL
means "less [sic] than 10 jobs") in 2009, and withheld values instead of zero for 2008-2014. Note: Delaware in 2007 is the only instance of theL
, so we might be able to get away with just subtracting 10 from 177 and calling it a day.The text was updated successfully, but these errors were encountered: