The study of student loan debt remains a timely and relevant higher education finance research and policy-oriented topic, especially when considering the alarming growth rates of student loan debt balances. The Quarterly Report on Household Debt and Credit released in May of 2018 shows that among all debt balances, student loans remain the only form of debt that virtually sextupled over the last 15-years, and this trend is not slowing down. Although aggregated trends are important, by definition they are limited in their capabilities to providing researchers, policy- and decision-makers with insights related to individual debt accumulation and, perhaps more importantly, with knowledge about the factors associated with variation of individual debt burden. Accordingly, the overarching goal of this study is to ameliorate this limitation in three meaningful ways. First, this is the first study that offers inferential estimates of the magnitude of student debt accumulation increase across two different decades (1991–2013) and institutional sectors (public 2- and 4-year colleges). Second, these estimates are based on student level undergraduate non-self-reported longitudinal loan debt disbursements. Third, the estimates not only account for individuals' baseline differences at the moment of college entry, but also account for institution- and state-level indicators that took place during college enrollment and that may be related to the variation of student loan debt reliance. Two nationally representative samples (NELS and ELS) complemented with other institution- and state-level data were analyzed using doubly robust estimators build from propensity score weights and entropy balancing approaches that were robust to unobservable selection issues using Oster's approach (J Bus Econ Stat 37(2):1–18, 2017). The results consistently indicated that, among all participants, student borrowing participation increased by 15 percentage points in the 2000s, compared to the 1990s, and individual debt accumulation at least doubled across decades. Notably, among 4-year degree holders, the 2-year path toward a 4-year degree consistently resulted in about 10% lower debt accumulation compared to the 4-year path toward a 4-year degree. Students who did not attain a 4-year degree were better served by having started college in the 2-year sector. In terms of overall debt increase, 4-year degree holders accrued about $8000 more on average than their counterparts did during the 1990s, however, the recent cohort also repaid about $11,000 more, on average (or three times as much), than participants did in the 1990s. These higher repayment behaviors observed among 4-year degree holders, resulted in similar amounts of their respective debt balances across decades. The implications are clear: students with higher propensities toward a 4-year degree attainment are likely to incur lower debt if they start college in the community college sector. However, before fully recommending this pathway, 2- and 4-year colleges' articulation agreements should be strengthened to ease transfer and eventual degree completion. Without recommending consolidation or merger between 2- and 4-year institutions, researchers and policy makers can learn from the strategies implemented by successful cases such as Perimeter College and Georgia State. Finally, 4-year entrants with lower likelihood to attain a 4-year degree may be better served by beginning college in the 2-year sector instead. Predictive analytics and machine learning techniques can be used to identify these cases, as depicted in the discussion section of the study. [ABSTRACT FROM AUTHOR]