The horizontal axis of the graph represents the students’ family income expressed as a percentage of the particular state’s median family income based on the State Median Family Income Estimates prepared by the U.S. Department of Health and Human Services Administration for Children and Families and published annually in the Federal register (e.g., Administration for Children and Families 2016).  We use the median income for the state and calendar year corresponding to the end of the academic year for which the graph is being constructed.  For example, if we are constructing a graph for the research institutions in the state of Washington in the 2011-2012 academic year we use the median family income for a family of four in Washington in 2012.

The vertical axis of the graph represents to funding that is devoted to paying for a student’s education for one academic year. The total amount spent by an individual student is the sum of the tuition & mandatory fees and the non-tuition costs including room & board, books & supplies, transportation, and personal expenses.

The total cost of attendance (COA) is displayed at the top of the vertical axis. Since the graph is constructed for a collection of institutions the value for the total cost of instruction that is displayed is a weighted average, weighted by full time equivalent (fte) in-state undergraduate enrollment, for the institutions in the collection. Tuition and mandatory fees are depicted with a horizontal line that is usually about centered top to bottom on the graph.  Note that the value of zero appears at the TOP of the dark gray band that extends across the bottom of the graph.

Enrollment and tuition & fees are needed to estimate the average COA and we determine them in one of two ways:  If we have “custom” data provided for a given state we use it, if we don’t have custom data we use data from the National Center for Education Statistics Integrated Postsecondary Education Data System (IPEDS).

Estimating the non-tuition & fees portion of the COA, i.e., cost of living and outfitting for school, is more challenging.  In the state of Washington the Washington Financial Aid Association (WFAA) conducts a periodic survey of students and generates model budgets annually (e.g., Washington Financial Aid Association 2016) that are used by campus financial aid administrators to develop student budgets for their individual campuses. Individual institutions report the non-tuition portion to the IPEDS.  We currently use the WFAA data for institutions within Washington and the IPEDS data almost everywhere else.


The state appropriation is is the funding for operations provided to institutions by the state.  It is included in the graph for reference purposes and it always appears as a dark gray band that extends across the bottom.  We attempt display the state appropriation on a per in-state fte undergraduate student.  For some states we have “custom” data but for most we use data reported by the institutions to the IPEDS.  Since the state appropriation is not a cost associated with individual students or their families, and is not included in the total cost of attendance, we consider the value of zero for the vertical axis as being located at the top of the gray band.


We estimate college savings by assuming that a family will invest some fixed portion of their discretionary income each year for some number of years.  We estimate discretionary income as the portion of gross income that exceeds some multiple of (or fraction of) the poverty guideline.  We use the poverty guideline that is published annually in the Federal Register by the U.S. Department of Health & Human Services Office of the Assistant Secretary for Planning and Evaluation (ASPE 2017) and we allow the user to determine the cutoff using a slider labeled “Family Income Exclusion.”  The product of the poverty guideline determines the point at which the dark green region that represents college savings becomes equal to zero on the graph. This point is indicated by the yellow circle in the illustration to the right. the yellow circle.

The user also sets the portion of discretionary income that is saved each year and the number of years for which savings occur using the sliders labeled “Percent disc income saved” and “Years of saving” respectively.  The user can also set an interest rate (real interest not APR since we use all current year values).  We calculate an end of series final value and then assume the accumulated savings are disbursed uniformly over the years that the student is attending college.


We estimate the family contribution using an approach similar to the approach used to estimate the annual amount saved in the college savings estimation described above.  The user is allowed to specify the percentage of the family’s discretionary income that is devoted each year to supporting the student in that year.


The Pell Grant award amounts for real students are determined using the Pell Award Schedule tables published annually by the Education Department (e.g., Department of Education 2016b).  The tables requires values for COA (see “The Framework for Constructing the Graph” above) based on the institution the student attends and the estimated family contribution (EFC) that is determined using the Expected Family Contribution Formula Guide (e.g., Department of Education 2016a) along with data describing family income, wealth, and obligations as determined by completing the Free Application for Federal Student Aid (e.g., Department of Education 2016c).

For our estimate of the Pell Grant award (dark blue region on the illustration to the right), we query the Pell Award Schedule table for the year being used to construct the graph. To do so, we use our estimate of the average COA for the institutions used to construct the graph (again please see “The Framework for Constructing the Graph” above) and we an estimate the EFC using a linear function of disposable income based on an exclusion at 200% of the Poverty Guideline.


The depiction of state funded financial aid is based on state policy.  We attempt to show the maximum or target award amount described in the policy.  To date we have state funded financial aid modeled for Washington, Mississippi, Minnesota, Oregon and Texas.


We use actual values for institutional aid provided by our partners in cooperating states.   Because these data are not always collected at a granularity that allows for a meaningful display across student family income levels, the only institutional aid data that is displayable at this time is from Washington and Texas.  We are currently working to improve this.  We intend to add an ability for users to add their own estimates for this function in the very near future.


We estimate the annual contribution by the student from their own current employment by assuming devotion of the total take home pay from a number of hours worked at minimum wage. The number hours devoted is selected by the user with a slider labeled “hours worked.” We use minimum wage information for individual states and years that is published by the U.S. Department of Labor (DOL 2016) to determine gross pay for those hours and then estimate the take home portion (net pay) using a ratio that was derived with an online paycheck calculator (ADP 2017).


We define affordable debt as the total debt for which the terms of repayment can be met by devoting a sum of money that is less than or equal to a specified portion, expressed as a ratio R, of the debtor’s annual earnings.  We call R the affordable debt ratio.  The affordable payment is P, where P = A x R   For example, in a case where a borrower’s annual earnings, A, are $36,000/yr and R is 10% the debtor could afford to spend P = 0.10 x $36,000/yr =  $3,600/yr (therefore $400/month) servicing the debt.  The total debt, D, that can be deemed “affordable” for that borrower would be that amount that resulted in payments that did not exceed $400/month.

The affordable debt (pink band) on the affordability graph is amount of that can be borrowed every year when the student is attending college, thus D/Y (where Y is the total years in college).

To determine D for someone with annual earnings of A we assume values for the loan terms i (interest, APR, on the borrowed funds) and n (number of equal monthly payments to repay the borrowed funds) and use the inverse of the common equation for determining monthly payments on a loan.  The program user can specify the loan terms i and n (the program default values are set to 5% and 120 payments respectively).

To determine a value of annual earnings, A, to in turn estimate affordable debt we use <<census bureau data and citation>> that represents the annual earnings for 25 to 34 years old individuals residing in the particular state, and holding the college degree expected, for the state and institutions for which the graph is being constructed.  The data represents the income distribution, we exclude the very low earners (people earning less than $5000/yr) and allow the user to select the percentile of earners after the exclusion is applied. Thus selecting the 50th percentile will result in an estimate of affordable debt for the median earner among the population being considered (e.g. 25-34 year old people with bachelors degrees who are earning at least $5000 per year in the state of YY).

In summary, to determine the estimate of affordable debt to display on the affordability graph we estimate A as described above and then use it along with the affordable debt ratio, R, to determine the maximum affordable annual payment, P.  Having determined P we use it along with the user selected (or default) values for the loan terms, i and n, to determine the maximum total affordable debt, D.  We plot the quantity D/Y on the affordability graph to represent the estimate of the amount of annual borrowing that can be deemed affordable.