I. INTRODUCTION

This page includes all the Compilations of my generosity comparisons for the 130 or so colleges analyzed in this website for the five academic years back from 2017/2018 through 2013/2014. My Compilations for the most current year, 2018/2019 are included in my Results page. I list a number of caveats in Part I of that page that apply here as well, so be sure to read it.

Additionally, please note that, negative "Remaining Balances" typically occur when highly generous schools have an expectation of student earnings that is lower than the $5,000 annual amount we use as a guide. Negative Remaining Balances are not refunded to students or their parents.

This page includes all the Compilations of my generosity comparisons for the 130 or so colleges analyzed in this website for the five academic years back from 2017/2018 through 2013/2014. My Compilations for the most current year, 2018/2019 are included in my Results page. I list a number of caveats in Part I of that page that apply here as well, so be sure to read it.

Additionally, please note that, negative "Remaining Balances" typically occur when highly generous schools have an expectation of student earnings that is lower than the $5,000 annual amount we use as a guide. Negative Remaining Balances are not refunded to students or their parents.

II. 2017/2018 COMPILATIONS

A. SOLVING FOR THE REMAINING BALANCE

Here is the Compilation for our $60K data set solving for the Remaining Balance:

Here is the Compilation for our $60K data set solving for the Remaining Balance:

acg-2017-_60k-compilation.pdf | |

File Size: | 428 kb |

File Type: |

And here is the Compilation for our $40K data set solving for the Remaining Balance:

acg-2017-_40k-compilation.pdf | |

File Size: | 428 kb |

File Type: |

B. SOLVING FOR EXPECTED ANNUAL LOANS

*Where the Remaining Balance minus the Expected Family Contribution equals the Expected Annual Loans.*

Here is the Compilation for our $60K data set solving for Expected Annual Loans:

Here is the Compilation for our $60K data set solving for Expected Annual Loans:

acg-2017-_60k-expected_loan_compilation.pdf | |

File Size: | 417 kb |

File Type: |

And here is the Compilation for our $40K data set solving for Expected Annual Loans:

acg-2017-_40k-expected_loan_compilation.pdf | |

File Size: | 459 kb |

File Type: |

III. 2016/2017 COMPILATIONS

A. SOLVING FOR THE REMAINING BALANCE

Here is the Compilation for our $60K data set solving for the Remaining Balance:

Here is the Compilation for our $60K data set solving for the Remaining Balance:

2016-_60k-compilation.pdf | |

File Size: | 437 kb |

File Type: |

And here is the Compilation for our $40K data set solving for the Remaining Balance:

2016-_40k-compilation.pdf | |

File Size: | 437 kb |

File Type: |

B. SOLVING FOR EXPECTED ANNUAL LOANS

*Where the Remaining Balance minus the Expected Family Contribution equals the Expected Annual Loans.*

Here is the Compilation for our $60K data set solving for Expected Annual Loans:

Here is the Compilation for our $60K data set solving for Expected Annual Loans:

2016-_60k-expected_loan_compilation-2.pdf | |

File Size: | 418 kb |

File Type: |

And here is the Compilation for our $40K data set solving for Expected Annual Loans:

2016-_40k-expected_loan_compilation-2.pdf | |

File Size: | 417 kb |

File Type: |

C. SOLVING FOR EXPECTED ANNUAL LOAN DIFFERENTIALS

The third Compilation in my loan series includes data on the differential between annual loan amounts for our middle and lower income sample families, and I want to offer a few preliminary comments before you review it.

As I put together both of the Expected Annual Loan Compilations, I noted that the quality of the news they represented varied a lot. The grant aid at some schools was generous enough to result in no expected loans for either sample family, which was wonderful. Loans were expected at the majority of the schools analyzed, not an unexpected result, but one fact took me by surprise. The expected annual loan amounts predicted by the NPC's at a noticeably large proportion of those schools were significantly higher for lower income students than they were for middle income students. As I analyzed all the data on loan differentials, I noticed that they naturally divided into four distinct categories with each interesting and important in its own way, and let's go through them one by one.

The amount of the Remaining Balance was less than or equal to the Expected Family Contribution at 29 of the 126 schools analyzed, representing 23% of the total. That meant that no loans were expected for either sample family at those schools each year, and that resulted in no loan amount differential. Loans were expected at the remaining 97 schools analyzed.

At 20 of the 126 schools analyzed, representing about 16% of the total, expected loan amounts for the lower income sample family were less than for the middle income sample family. That appears to be a gratifying result, and I am not the guy to argue against increased college generosity, but that advantage for lower income families isn't really logical. Obviously, students from lower income families should not be penalized with higher loans than middle income students, but the ability to repay a student loan is dependent on the income of the students, not their parents. So, I think that an equivalence of loan amounts for both sample families is the most logical result, but I can see why colleges might want to err on the generous side to avoid the potential problems discussed below.

31 of the 126 schools analyzed, about 25%, had loan amount differentials which were functionally equal to zero, and I find this the most interesting segment of this Compilation. Tracking big breaks in the data and realizing that zero might be a difficult outcome to reliably hit, I defined "functionally equal to zero" as values between +$250 and -$250 per year. This meant that this segment included 501 potential outcomes, not forgetting zero as a potential outcome. Being very used to the even spray of outcomes in my other Compilations over the last four years , I expected these 31 outcomes to include the whole 501 unit range with no - or almost no - duplication, but that was not the case at all. Only four numeric outcomes covered 23 of the 31 schools, meaning 74% of the segment; and there were only eight solo outcomes. It took me a while to figure it out, and this is my best guess at what's going on here: all 31 schools with functionally zero loan differentials are doing their best to approximate that worthy goal; to do this, 8 schools are using their own algorithms, and 23 schools are using publicly available algorithms. Take a look at the data and think about the vastly different schools achieving exactly the same results, and see if you don't agree with me. I mean, when you've got 13 schools at the plus $18 outcome, and those schools range from Sweet Briar to Berkeley and UCLA, I think they have to be using the same algorithm.

Sadly, the last segment in this Compilation is also the largest. At 46 of the 126 schools analyzed, about 37% of the total, expected annual loans as predicted using the school's own Net Price Calculator programs for the sample family with an AGI of $40,000 per year were more than the expected annual loans for the sample family with an AGI of $60,000 per year. The dollar differential ran from $527 to $5,872 in additional expected loans per year. I find that ethically and morally unacceptable, and I think most Americans would agree. Additionally, the potential practical consequences of this differential for these schools are vast. The list of identified schools includes a large number of otherwise admirable institutions, my own*alma mater *is one of them, and they should all fix this problem.

The third Compilation in my loan series includes data on the differential between annual loan amounts for our middle and lower income sample families, and I want to offer a few preliminary comments before you review it.

As I put together both of the Expected Annual Loan Compilations, I noted that the quality of the news they represented varied a lot. The grant aid at some schools was generous enough to result in no expected loans for either sample family, which was wonderful. Loans were expected at the majority of the schools analyzed, not an unexpected result, but one fact took me by surprise. The expected annual loan amounts predicted by the NPC's at a noticeably large proportion of those schools were significantly higher for lower income students than they were for middle income students. As I analyzed all the data on loan differentials, I noticed that they naturally divided into four distinct categories with each interesting and important in its own way, and let's go through them one by one.

The amount of the Remaining Balance was less than or equal to the Expected Family Contribution at 29 of the 126 schools analyzed, representing 23% of the total. That meant that no loans were expected for either sample family at those schools each year, and that resulted in no loan amount differential. Loans were expected at the remaining 97 schools analyzed.

At 20 of the 126 schools analyzed, representing about 16% of the total, expected loan amounts for the lower income sample family were less than for the middle income sample family. That appears to be a gratifying result, and I am not the guy to argue against increased college generosity, but that advantage for lower income families isn't really logical. Obviously, students from lower income families should not be penalized with higher loans than middle income students, but the ability to repay a student loan is dependent on the income of the students, not their parents. So, I think that an equivalence of loan amounts for both sample families is the most logical result, but I can see why colleges might want to err on the generous side to avoid the potential problems discussed below.

31 of the 126 schools analyzed, about 25%, had loan amount differentials which were functionally equal to zero, and I find this the most interesting segment of this Compilation. Tracking big breaks in the data and realizing that zero might be a difficult outcome to reliably hit, I defined "functionally equal to zero" as values between +$250 and -$250 per year. This meant that this segment included 501 potential outcomes, not forgetting zero as a potential outcome. Being very used to the even spray of outcomes in my other Compilations over the last four years , I expected these 31 outcomes to include the whole 501 unit range with no - or almost no - duplication, but that was not the case at all. Only four numeric outcomes covered 23 of the 31 schools, meaning 74% of the segment; and there were only eight solo outcomes. It took me a while to figure it out, and this is my best guess at what's going on here: all 31 schools with functionally zero loan differentials are doing their best to approximate that worthy goal; to do this, 8 schools are using their own algorithms, and 23 schools are using publicly available algorithms. Take a look at the data and think about the vastly different schools achieving exactly the same results, and see if you don't agree with me. I mean, when you've got 13 schools at the plus $18 outcome, and those schools range from Sweet Briar to Berkeley and UCLA, I think they have to be using the same algorithm.

Sadly, the last segment in this Compilation is also the largest. At 46 of the 126 schools analyzed, about 37% of the total, expected annual loans as predicted using the school's own Net Price Calculator programs for the sample family with an AGI of $40,000 per year were more than the expected annual loans for the sample family with an AGI of $60,000 per year. The dollar differential ran from $527 to $5,872 in additional expected loans per year. I find that ethically and morally unacceptable, and I think most Americans would agree. Additionally, the potential practical consequences of this differential for these schools are vast. The list of identified schools includes a large number of otherwise admirable institutions, my own

2016_compilation_of_expected_annual_loan_differentials-2.pdf | |

File Size: | 408 kb |

File Type: |

IV. 2015/2016 COMPILATIONS

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

2015-_60k-compilation.pdf | |

File Size: | 433 kb |

File Type: |

B. And here is the Compilation for our $40K data set solving for the Remaining Balance:

2015-_40k-compilation.pdf | |

File Size: | 435 kb |

File Type: |

V. 2014/2015 COMPILATIONS

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

2014-_60k-compilation-6.pdf | |

File Size: | 384 kb |

File Type: |

B. And here is the Compilation for our $40K data set solving for the Remaining Balance:

2014-_40k-compilation-3.pdf | |

File Size: | 438 kb |

File Type: |

VI. 2013/2014 COMPILATIONS

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

A. Here is the Compilation for our $60K data set solving for the Remaining Balance:

2013__60k_compilation.pdf | |

File Size: | 344 kb |

File Type: |

B. And here is the Compilation for our $40K data set solving for the Remaining Balance:

2013__40k_compilation.pdf | |

File Size: | 330 kb |

File Type: |

Copyright 2019, Mark Warns, All Rights Reserved

Here again are a Word version of my Apple-to-Apples template for your own college cost comparisons along with a Word version of my blank data input pages that you can download for use by your own family:

acg_2018-2019_apples-to-apples_template.docx | |

File Size: | 17 kb |

File Type: | docx |

acg_2018-2019_family_npc_input_data.docx | |

File Size: | 19 kb |

File Type: | docx |

Proudly powered by Weebly