Transparency in ISA Program Design

Jul 15, 2021

Design Tool 1.0

It has been a little over a year since Leif made its Program Design Tool available to the public (see blog post) in addition to the hands-on consultative services we provide our partner schools. The modeling framework can be used as a valuable tool for schools seeking to launch an Income Share Agreement (ISA) program. The goal was to power a thoughtful approach to ISA program design across the industry. The approach ensures schools are incentivized to deliver positive outcomes while also providing the necessary protections to students.

Over the past year we have seen countless schools leverage the tool to explore the creation and launch of successful ISA programs. We have also seen the benefits of making these resources available to the public so that all schools can start the journey of designing sustainable, incentive-aligned tuition payments solutions.

Today we would like to dive deeper into what we have learned from schools using the tool, its primary merits, and observations gained along the way.


As previously highlighted, Design Tool 1.0 had several key Leif-imposed guardrails put in place for the benefit of students. Several of these are listed below 1:

  • Payment Cap
  • Income Share Percentage
  • Minimum Income Threshold
  • Affordability Limits

1 Please note that such guardrails are applied at the portfolio level and are not determined on an individual basis.

Real-life View of Outcomes

Instead of linearly estimating payback using the expected median income and placement rate of a program, Leif uses those data points to respectively model both expected salary and placement distributions.

The end result allows Leif's modeling framework to estimate a realistic distribution of expected monthly payments and expected contract paybacks for a hypothetical student. This ultimately allows us to ensure that affordability considerations hold true for students both at the high-end and low-end of said distributions.

Furthermore, by accounting for several factors — time spent in the program, expected duration of the job search, and the timing of upfront payments as an alternative — we model out reasonable program terms while taking affordability into consideration. This also allows us to disregard scenarios where the implied costs to students deviate greatly from alternative forms of payment for the program.

Affordability Controls

Concentration Limits

By drawing from the expected income distributions, Leif is able to model the share of students that are expected to pay a specific multiple of the upfront cost of the program. Leif's modeling framework imposes a hard cap on how many students are expected to pay said multiples, thus eliminating program terms that rely heavily on high-earner subsidization.

Take-home Pay Limits

Leif's modeling framework analyzes the take-home pay (gross salary net of ISA payment) to disqualify program terms that may place the students at the low end of the income distribution below reasonable income limits. This ensures proposed terms do not inadvertently cause undue burden on lower earning students.

Efficient Minimum Income Thresholds

Leif's modeling framework ensures lower income earners enjoy the downside protection of minimum income thresholds that are unique to ISAs. We set the floor on the minimum income threshold at the highest of (a) the first decile of the expected income distribution for the field of study, or (b) an absolute minimum threshold of $25k, regardless of the field of study.

Program Sustainability

To ensure schools can operate programs sustainably our design engine emphasizes terms that result in average paybacks that generally track the cost of schools to train their students. Without the right consultative services from an established program manager, poorly designed programs can be destined for failure before they even launch.


Income Distribution and Non-qualified Placements

As noted before, utilizing income distributions paints a more realistic picture of the expected payback across a cohort of ISAs. That said, depending on the shape of the income distribution, two programs with the exact same median income and placement rates might have different expected payback under the same ISA terms.

Since the initial version of the tool was designed to generally track the cost of schools to train their students, this sometimes resulted in ISA terms that relied more heavily on subsidization from high-earners. This was an accurate representation of the expected cash flows and simply the mathematical output of the tool's chosen methodology (see Example #1 below).

Example #1:

This example highlights the subsidy allocation due to the fact that the wider income distribution for the first profession results in a higher subsidization coming from high-earners.

Scenario A: Music Studies [wider income distribution]


Scenario B: Engineering [narrower income distribution]


Placement Rates and Subsidy Allocation

Another observation from the initial version of the design tool involves the subsidization required to operate programs with particularly low placement rates. The framework allocated the subsidization to schools from lower expected payments at the cohort level but also to students through higher lifetime ISA payback at the individual level (see Example #2 below).

This was an unintentional effect of not imposing strong enough constraints on the source of the subsidies. When schools have low historical placement rates it is indeed true that students will derive greater benefit from the downside wage insurance provided by ISAs. That said, we believe that schools should bear the additional cost of programs that underperform reasonable benchmarks.

Example #2:

This example highlights the subsidy allocation across both schools (lower Average Payback) and students (higher Average Qualified Payback)

Scenario A: High Placement Rate


Scenario B: Low Placement Rate


Design Tool 2.0

With a constant focus on improving outcomes for our partner schools, Leif now enforces a fixed placement rate of 80% as an anchor in our term generation design tool. This feature holds our partner schools to high standards of outcomes while ensuring that the actual school placement rate dictates the expected payback that the school will realize under the recommended set of terms. Ultimately this fully allocates the subsidization to the school's expected payback. Furthermore, our improved tool no longer emphasizes high-earner subsidization to compensate for income distribution variability.

These are important improvements as they effectively standardize the term recommendations to an objective standard of placement outcomes. The two examples below (Example #3 and Example #4) provide deeper insight into the most recent version of the Design Tool and specific improvements on program design for the two previous examples provided earlier in the post.

Example #3: Different Profession, Same Median Outcomes

This example highlights the improved subsidy allocation towards schools (lower Average Payback) while holding students' Average Qualified Payback constant

Scenario A: Music Studies [wider income distribution]


Scenario B: Engineering [narrower income distribution]


Example #4: Same Profession, Different Placement Rates

This example highlights the improved subsidy allocation towards schools (lower Average Payback) while holding students' Average Qualified Payback constant.

Scenario A: High Placement Rate


Scenario B: Low Placement Rate


Aligned Incentives Lead to Positive Outcomes

Leif seeks to increase access to quality and affordable education by a consistent focus on improving student outcomes. We are happy to provide our modeling framework to the public and believe a better understanding of thoughtful program design is a constructive exercise for the entire education finance industry. The increased transparency will only serve to highlight the inherent student protections that are unique to the asset class — such as downside wage protection. More than ever an alignment of incentives between schools and students is required to ensure positive outcomes in the education industry.