This is the third piece of a four-part series about data literacy and critical data consumerism through the lens of economic prosperity. You can read Part 1: Income and Part 2: Unemployment here.

In this series, we explore four variables associated with economic prosperity through the lens of different data sources: Income, unemployment, student loan payments, and mortgage interest payments. As you will see, each data source paints a slightly different picture of the state depending on how the question was asked, the data source, and the respondent population. We hope that this series will remind you that we must all continue our curiosity about the information presented to us with and work to be critical consumers of data in our everyday lives.

The analyses in this blog post were completed by our Fall 2019 Wesleyan Intern, Spencer Arnold, and CTData Data Analyst, Jason Cheung.

Part 3: Student Loans

Research student loans and the term “crisis” tends to appear in the search results. Student loan debt has been increasing steadily as the cost of education rises in our country, with student loan debt accounting for over $1.5 trillion in 2019 compared to $260 billion in 2004. In fact, 11% of consumer debt is attributed to student loans with the average student loan debt ringing in at around $37,000.

Data Sources

The Institute for College Access and Success

The Institute for College Access and Success provides an online repository of data and research related to higher education called College Insight. Topics include student loan debt, college affordability, and student success.

Federal Reserve

The Federal Reserve has conducted the Survey on Household Economics and Decisionmaking (SHED) since 2013 with approximately 12,000 individuals each year. SHED provides data about the financial status of adults in the United States including education, student loans, employment, dealing with unexpected expenses, credit and banking, housing, and retirement.

IRS Form 1040

The U.S. Department of Internal Revenue Services collects annual income tax returns from taxpayers using Form 1040. IRS Form 1040 also captures deductions for government assistance programs such as unemployment. The maximum student loan deduction that can be claimed is $2,500 and only those making below $200,000 are eligible for this deduction.

2017 At-a-Glance

The Institute for College Access and Success (TICAS) found that the Connecticut graduating class of 2017 had the highest education debt in the nation at $38,500. This is nearly $10,000 more than the national average of $29,650 for graduating senior borrowers from four-year colleges. Seventeen other states had an average student loan debt over $30,000, which included many of Connecticut’s New England neighbors such as Rhode Island ($36,250), New Hampshire ($34,415), Massachusetts ($32,065), and Maine ($31,364).

Contrastingly, the Federal Reserve’s 2017 SHED survey found a lower average than TICAS, with the average outstanding student loan debt for borrowers being $20,000 to $25,000. Additionally, the survey found that almost one third (30%) of adults who attended college had student loan debt. For those with debt, approximately one-fifth were behind on their loan payments. In addition to student loan debt, SHED found that students used other sources to finance their education: 25% borrowed with credit cards, 6% took out a home equity line of credit, and 7% used another form of loan. This data was only available nationally, not disaggregated for individual states.

IRS individual tax return data paints yet another picture of student loans. The average student loan deduction for taxpayers in Connecticut was $1,124. At approximately 9% of taxpayers, Connecticut residents ranked 22nd in the country for the percentage who made a payment on a student loan in 2017. When looking at these numbers by income, we see that lower income taxpayers in Connecticut claimed higher student loan deductions, as compared to the United States average.

2007-2017 Trends

Looking longitudinally at The Institute for College Access and Success, we found that between 2007 and 2017, college graduates in Connecticut consistently had higher student loan debt than the national average. This gap has increasingly widened since 2014 (see chart below). The Federal Reserve reported a similar trend nationally with the average student loan debt in the United States ranging from $25,750 (median: $13,000) in 2013 to $32,731 (median: $17,000) in 2016. The data was reported differently in 2017 and not available for comparison over time. College graduates in Connecticut saw a 73% increase in student loan debt between 2007 (US: $863, CT: $901) and 2017 (US: $1,083, CT: $1,124) as compared to a 41% increase nationally. According to IRS individual tax return data, Connecticut taxpayers with student loan payments were consistently claiming higher deductions than the US average between 2007 and 2017. This was likely attributed to the fact that Connecticut graduates have higher student loan debt than the national average.

Furthermore, findings from the Federal Reserve’s SHED survey suggest that student loan debt may be impacting the financial stability and wealth accumulation of students of color differently than white students. As shown in the chart below from the Federal Reserve’s Report on the Economic Well-Being of U.S. Households in 2017, a smaller percentage of students of color have paid off their student loans as compared to white students.

Source: The Federal Reserve, Report on the Economic Well-Being of U.S. Households in 2017-May 2018

Why are there differences in reported student loan data?

Similar to Parts 1 and 2 of this series, it is evident that the differences in student loan data between the three sources are not necessarily due to errors in the data but rather how the data is collected and reported. For example, the three data sources used:

Different data collection methods
- TICAS is self-report from universities that participated.
- SHED is a survey, which means that the data is self-reported and unverified.
- IRS Form 2014 is verified and audited by the government.
Different samples and exclusionary criteria are included in each of the sources
- TICAS only included data on students in the most recent graduating class who attended public or nonprofit universities. Because TICAS allowed universities to opt into the sample, results represented data from approximately half of the eligible universities.
- The Federal Reserve’s SHED survey was administered with a representative probability-based sample selected by address-based sampling. Requirements included being over age 18 and living in the United States, with those making under $40,000 per year being oversampled. There was a monetary incentive to complete the survey.
- Student loan data from IRS Form 1040 excluded individuals making over $200,000 a year, since those individuals are ineligible to deduct student loan interest payments.

So What?

Taking out loans to pay for education is not inherently bad. It can help grow a student’s credit, invest in their future by helping to build a career, and teach important financial literacy skills. The challenge is as the cost of education rises, many students are amassing large amounts of debt in pursuit of careers that may not have salary structures conducive to paying off their loans. Others may acquire student loan debt and leave school prior to receiving their degree. As the length to pay off student loans increases, students are faced with difficult choices such as delaying buying a home or starting a family, or choosing between paying off their debt and affording basic needs such as food, housing, and medical costs.

We can’t immediately trust every data point we find. By looking at differences in data sources, collection, and reporting, we can use data literacy to assess information and integrate various perspectives to help us develop a more holistic view of an issue. If you’d like to read more from our Source Shapes the Story series, check out Part I: Income and Part II: Unemployment. For tips on how to be a more critical data consumer, head to the Data Literacy section of our blog or attend one of our virtual CTData Academy workshops. You can also stay informed by signing up for our newsletters and following us on Twitter, Facebook, and LinkedIn.

Income in Connecticut: The Source Shapes the Story—Part III, Student Loans