Closed andersonfrailey closed 6 years ago
@Kpinkelman, you can tabulate a high and low fraction of in-kind benefits in other benefits by doing this:
High fraction is computed interpreting Both
and <blank>
as In-Kind
, and
Low fraction is computed interpreting Both
and <blank>
as Cash
.
In both cases be sure to omit unemployment benefits.
Can you post the results of those tabulations in this issue? Thanks.
@andersonfrailey
@martinholmer , I’m a bit confused as to what you mean by this – My understanding of what you’re asking:
In researching which of these programs offered cash vs in-kind benefits, I looked into how they listed their benefits online. Archives containing information about outlays didn’t detail where the money itself went (either directly to individuals or paying for services).
To calculate High fraction and Low fraction, where should I draw these numbers from?
@martinholmer said:
you can tabulate a high and low fraction of in-kind benefits in other benefits by doing this:
High fraction is computed interpreting Both and
as In-Kind, and Low fraction is computed interpreting Both and
as Cash. In both cases be sure to omit unemployment benefits.
Sorry if this wasn't clear. All I need to know is the range of value of the in-kind fraction for other_benefits
.
If you had classified each one as either In-Kind
or Cash
it would be simple to add up the dollars amounts of all the benefits (other than unemployment compensation) that are labeled In-Kind
and divide it by the total of all benefits (again omitting unemployment compensation). Does that make sense?
But above is not the case. You have two categories that are blank and several others that are labeled Both
.
All I'm asking is that you tabulate two in-kind fraction estimates. The first (high) estimate assumes that all the benefits that are blank or Both
are assumed to be In-Kind
. The second (low) estimate assumes that all the benefits that are blank or Both
are assumed to be Cash
. Does that make sense?
@Kpinkelman , Is this explanation more clear?
@Kpinkelman asked:
To calculate High fraction and Low fraction, where should I draw these numbers from?
From the table shown in the first comment in this issue. Didn't you create that table?
@martinholmer, That makes more sense to me. Results below:
Type | Estimate |
---|---|
Low Fraction (In-Kind/Total) | .52163293 |
High Fraction (In-Kind/Total) | .76352877 |
@Kpinkelman, Thanks so much for the range tabulation here.
If your research can narrow the range that would be great.
@andersonfrailey
@Kpinkelman a couple questions:
Do you have the source(s) for this table handy?
Is Family support payments to States and TANF (currently classified as cash) only the cash component of TANF? As of 2014, TANF was ~25% cash, plus another ~10% tax credits (per 538), the rest being in-kind. With the $20.4B cost and 2014 TANF budget of $17.4B I'm guessing it may include all of TANF, though I'm not familiar with the family support payments.
Is there a rough list of the types of programs included in the two "Other" rows?
@MaxGhenis sorry for the delayed response:
The numbers in the table had already been filled out when I received it, but are from budget docs released by the Obama administration (namely, 3.2 and 11.3 on https://obamawhitehouse.archives.gov/omb/budget/Historicals). Unfortunately they don't contain information about how the benefits are administered (in-kind or cash), so that information was mostly gleaned from reading up on descriptions of the programs which likely explains 2.
You're likely right here - I'm working on a new version of this table and I'll make sure to change that and double check if any similar mistakes were made with other programs.
There isn't currently, but I can look into that and compile a list.
Updated Table:
Program | Table | Row | 2014_cost | In-Kind or Cash |
---|---|---|---|---|
Unemployment Assistance | 11.3 | 16 | 43504 | |
Children's health insurance | 11.3 | 20 | 9317 | In-Kind |
Indian health | 11.3 | 22 | 4510 | In-Kind |
Health resources and services | 11.3 | 24 | 7604 | In-Kind |
Substance abuse and mental health services | 11.3 | 25 | 3193 | In-Kind |
Center for Medicare and Medicaid Innovation | 11.3 | 27 | 997 | In-Kind |
Refundable Premium Tax Credit and Cost Sharing Reductions | 11.3 | 28 | 13068 | Cash |
Other - Medical | 11.3 | 30 | 12834 | |
Student assistance--Department of Education and other | 11.3 | 34 | 56337 | In-Kind |
Housing assistance | 11.3 | 36 | 46600 | In-Kind |
Child nutrition and special milk programs | 11.3 | 39 | 19490 | In-Kind |
Supplemental feeding programs (WIC and CSFP) | 11.3 | 40 | 6266 | Both |
Commodity donations and other | 11.3 | 41 | 823 | In-Kind |
Family support payments to States and TANF | 11.3 | 45 | 20378 | Both |
Low income home energy assistance | 11.3 | 46 | 3537 | In-Kind |
Payments to States for daycare assistance | 11.3 | 48 | 5064 | In-Kind |
Payments to States--Foster Care/Adoption Assist. | 11.3 | 50 | 6868 | Cash |
Other public assistance | 11.3 | 53 | 1071 | |
Coal miners and black lung benefits | 11.3 | 56 | 426 | Both |
Aging services programs | 11.3 | 58 | 1462 | Both |
Energy employees compensation fund | 11.3 | 59 | 1052 | Cash |
September 11th victim compensation | 11.3 | 60 | 49 | Cash |
Refugee assistance and other | 11.3 | 61 | 4403 | Both |
351 Farm income stabilization | 3.2 | 42 | 20012 | Cash |
352 Agricultural research and services | 3.2 | 43 | 4374 | In-Kind |
451 Community development | 3.2 | 62 | 7896 | In-Kind |
452 Area and regional development | 3.2 | 63 | 3027 | In-Kind |
453 Disaster relief and insurance | 3.2 | 64 | 9747 | Both |
506 Social services | 3.2 | 72 | 17299 | Both |
Potential Types of Programs in Others - Medical (likely more than this):
SAMHSA, Inmate healthcare, Federal Refugee Health Promotion Program, Federal Employees Health Benefits Program, Tricare, Healthy Start, etc
Potential Types of Programs in Others - Public Assistance (likely more than this):
Educational assistance, Pell Grants, Head Start, Job Training, Lifeline
@andersonfrailey, Is there still something to be discussed in issue #148? This issue was opened on February 7, 2018, and there has been no discussion in it or reference made to it since March 3, 2018. If the issue is resolved, you might want to closed it.
I believe we can close this issue now. If anyone involved has further questions they're free to reopen it.
This spreadsheet starts with @Kpinkelman's table from https://github.com/PSLmodels/taxdata/issues/148#issuecomment-367119322 and adds a column to exclude benefits that are counted outside other_ben
. I excluded the following four programs to avoid what I believe would be double-counting:
housing_ben
)e02300
)tanf_ben
)wic_ben
)Based on this, the cash share of other_ben
ranges from 17% to 35%, depending on if you consider programs labeled Both to be 0% or 100% cash, respectively.
Our new intern, @KPinkelman, did some research into which of the programs in
cps_data/benefitprograms.csv
offered cash benefits and which offered in-kind benefits. Here is a table of his results:And the spreadsheet file for those interested: program_types.xlsx
I told him to leave out Unemployment Assistance because we do not count it in the ultimate calculation of
other_ben
.A couple of questions remain for our work:
cc. @martinholmer