Read on Twitter
Ok, since #IncomeDrivenRepayment is a topic in the #CancelStudentDebt conversation - I have a few things to say, which I'll do in chronological order.
(1)
(1)
First, we do not have a very clear understanding of who may already be enrolled in IDR, and part of this issue is due to limited datasets that allow us access to the variables that would be important to understand this.
Check out @n_hillman and @elliebruecker piece.
(2)
Check out @n_hillman and @elliebruecker piece.
(2)
Earlier this year, I published my first paper on IDR from data I captured during my dissertation. This is a non-nationally represented sample but it has guided subsequent analyses.
https://ir.library.louisville.edu/jsfa/vol49/iss2/3/
(3)
https://ir.library.louisville.edu/jsfa/vol49/iss2/3/
(3)
In short, there were descriptive differences between those in IDR and those in traditional repayment
Those in IDR had more debt (duh) and unique financial outcomes - but they paid statistically the same amount per month for federal debt.
(4)
Those in IDR had more debt (duh) and unique financial outcomes - but they paid statistically the same amount per month for federal debt.
(4)
Regressions positively linked the following to IDR enrollment
Female borrowers
Masters Degrees
Professional Degrees (not robust)
Loan debt starting at $60k (Ref $20k or less)
All GAI below $75K (Ref $100k or more), earners $25-39k being most likely
Non-Metro Residency
(5)
Female borrowers
Masters Degrees
Professional Degrees (not robust)
Loan debt starting at $60k (Ref $20k or less)
All GAI below $75K (Ref $100k or more), earners $25-39k being most likely
Non-Metro Residency
(5)
The reference groups were chosen due to already established narratives surrounding IDR - like higher-earning people taking advantage of the policy.
(6)
(6)
I then tested to see if enrollment in IDR correlated with respondents' self-reported ability to put money into savings, retirement, and were homeowners.
A negative link existed between IDR and the ability to save but did not establish a significant link with the others.
(7)
A negative link existed between IDR and the ability to save but did not establish a significant link with the others.
(7)
With @FitzEdPolicy and @ChrisMarsicano - we then used the SCF '16 dataset to test my former models and generated models with binary +$50k models influenced by Looney & Yannelis’ (2018)...
https://preprints.apsanet.org/engage/apsa/article-details/5e8b3bedcf138e0019f49641
(8)
https://preprints.apsanet.org/engage/apsa/article-details/5e8b3bedcf138e0019f49641
(8)
RE: IDR enrollment, we found that depending on the model specification unique outputs were generated - which is not exactly what you'd like to see. As the models were not cemented we focused on the trends between the models and my former paper
(9)
(9)
So we concluded the following variables positively linked with enrollment
Female borrowers
Minority borrowers
Borrowers with $50k+
"Some College"
And consistently those earning <$12.5k were negatively correlated (more on this in a sec)
(10)
Female borrowers
Minority borrowers
Borrowers with $50k+
"Some College"
And consistently those earning <$12.5k were negatively correlated (more on this in a sec)
(10)
We also looked for whether being in IDR correlated to the outcome in the picture below. All of these were null.
We suggest that despite these individuals having more debt, IDR may be producing a measure of equalization.
(11)
We suggest that despite these individuals having more debt, IDR may be producing a measure of equalization.
(11)
Just last week, we re-ran our analyses using the SCF '19 dataset and found more information. Before I get into that, my first paper did not have REPAYE included nor did the SCF '16. The SCF '19 does...
(12)
(12)
So we retested my models and Looney & Yannelis’ (2018) inspired ones. AGAIN depending on how you specified the models we got unique outputs (damn it!).
With that said we looked for trends between the regressions and linked them to our prior work.
(13)
With that said we looked for trends between the regressions and linked them to our prior work.
(13)
Again, female + racial minority borrowers are more likely to enroll in IDR.
Those above median earnings are more likely but also those with $100k+ in wages were less likely. Wages beyond <$12.5 were not sig in '16 paper.
The lowest earners were again less likely
(14)
Those above median earnings are more likely but also those with $100k+ in wages were less likely. Wages beyond <$12.5 were not sig in '16 paper.
The lowest earners were again less likely
(14)
This time we found significant trends linking IDR enrollment with:
1. Checking amount (-$1,004)
2. Savings for Retirement (Y/N), -7pp
3. Retirement Amount (-$8,202)
So we are starting to see divergent trends...
(15)
1. Checking amount (-$1,004)
2. Savings for Retirement (Y/N), -7pp
3. Retirement Amount (-$8,202)
So we are starting to see divergent trends...
(15)
The SCF '19 trends require more focus - which we have planned; however, we believe that the inclusion of REPAYE has lowered the floor for entry at loan debt STARTING at $40k is now a significant access point. This was NOT found in my first paper or the '16 paper.
(16)
(16)
SO... what are we comfortable suggesting between the papers? Changing IDR policies would affect
1. Female Borrowers
2. Minority Borrowers
3. Those with higher debts
4. Median Earners (and slightly higher)
...... next Tweet, too
(17)
1. Female Borrowers
2. Minority Borrowers
3. Those with higher debts
4. Median Earners (and slightly higher)
...... next Tweet, too
(17)
And it may change some findings that suggest IDR is keeping certain "American Dream" outcomes statistically equal.
See more here
https://www.thirdway.org/report/income-driven-repayment-whos-using-it-and-how-covid-19-could-change-the-landscape
(18)
See more here
https://www.thirdway.org/report/income-driven-repayment-whos-using-it-and-how-covid-19-could-change-the-landscape
(18)
NOW - when it comes to people saying, "Well everyone should just be part of IDR" instead of debt cancelation straight out.
1st: that's going to cost the US Government in the long run when/if people get to forgiveness.
(19)
1st: that's going to cost the US Government in the long run when/if people get to forgiveness.
(19)
2nd: This may not help people pay down their debt but it would help ease how much they have to pay per month
3rd: Many (not all) of the same people saying "go to IDR" will also be the people who are "concerned" over the long-term costs to the U.S.
(20)
3rd: Many (not all) of the same people saying "go to IDR" will also be the people who are "concerned" over the long-term costs to the U.S.
(20)
4th: Encouraging IDR enrollment may be a fair suggestion as financial outcomes appear statistically equal but my work doesn't capture non-financial aspects (stress/health) we can't speak to these and they MUST BE a focus of future study before anyone says "they'll be fine."
(21)
(21)
I'll be honest with my beliefs, cancelation up to $50k would be one of the better ways to go because so many people's lives would be dramatically changed.
But before going with "just put them in IDR" maybe you should know a bit about what's been going on with IDR?
(22)
But before going with "just put them in IDR" maybe you should know a bit about what's been going on with IDR?
(22)
My belief is cemented in this data and the qual responses I have (but have done nothing with). The pain financial and non-financial people have expressed is quite outstanding and we need more people to convey this to the rest. I feel as if I have failed in this aspect.
(23)
(23)
That's it... that's all I had for now.
Next year, I'll have more things to say with our #SLLI grant and data.
(End)
Next year, I'll have more things to say with our #SLLI grant and data.
(End)
