Episode 2:
How we did the digging
Process, sources, and assumptions
To explore the relationship between college degrees, living-wage employment, and whether graduates stay in Hawaiʻi, we brought together two distinct datasets — each with strengths, gaps, and unique insight.
Wage Data
We began with graduate earnings data from the Hawaiʻi Data Exchange Partnership (DXP), which provides one-to-five-year institution- and program-level wage outcomes for University of Hawaiʻi graduates by campus and program. For this episode, we focused on:
Four-year degree earners (bachelor’s and post-baccalaureate certificates)
Graduates from academic years 2011–12 through 2015–16
Those found working in Hawaiʻi for 3 or 4 quarters of the year, earning at least minimum wage
This allowed us to chart earnings trajectories over time — from one to five years after graduation — by program and by campus.
Wage + Location Data
To understand where graduates end up, we drew on the Census Post-Secondary Employment Outcomes (PSEO) Explorer, a national dataset from the U.S. Census that shows where graduates are working (in or out of state) and median earnings at one, five, and ten years. This dataset shows:
Median earnings for graduates by field of study
The percentage of graduates still living in Hawaiʻi 1, 5, and 10 years after graduation
Unlike the DXP dataset, which is specific to UH programs, the PSEO data is grouped into standardized CIP categories (fields of study) that allow for national comparison but don’t reflect UH’s full program-level specificity. Still, it offers an important lens into geographic retention patterns — helping us see which degree fields are more likely to keep graduates local.
How We Combined the Data and what comes next
To build a fuller picture, we created a blended dataset that maps UH program-level wage data (from DXP) to field-of-study location outcomes (from PSEO). While the datasets don’t align perfectly — the cohorts cover different time periods, and some programs don’t map cleanly — the match is close enough to reveal meaningful patterns that neither dataset can show alone.
This composite approach allowed us to:
Visualize how much graduates earn, by degree and campus
Show how long it takes to reach a living wage
Plot what percentage of graduates are still in Hawaiʻi at years 5 and 10
Build a quadrant view of wage vs. location outcomes to explore which degrees lead to good jobs and support local staying power.
Together, these tools let us answer critical questions about higher education’s return on investment — especially for students trying to stay in Hawaiʻi. But they also highlight a major blind spot: what about the 50% of learners who never enroll in college? We don’t currently have the data to understand their employment journeys, wage trajectories, or long-term outcomes.
That’s where the work of the Data Sharing and Governance Working Group, established under Act 154, becomes essential. Act 154 created a formal working group within the Office of Enterprise Technology Services to tackle exactly this challenge: how to securely and strategically share data across state agencies — education, labor, human services, health, and more — so Hawaiʻi can:
Track outcomes for all residents, not just college-goers
Understand whether career pathways are working — and for whom
Modernize unemployment and wage record systems to include details like job title and location
Build public dashboards that empower learners, families, policymakers, and funders
Other states like California, Colorado, and Washington have already taken bold steps to centralize data coordination, improve interoperability, and develop governance models that protect privacy while unlocking insights. Act 154 is Hawaiʻi’s first step in that direction.
A Note on Limitations
This analysis reflects the best available data — but there are important limitations:
Different cohort years: DXP and PSEO cover slightly different graduating classes, which may affect comparability.
Program mapping: UH program names were matched to broader PSEO categories, introducing generalization.
Workforce visibility: Both datasets only include graduates found in administrative wage records (e.g., UI wage data), meaning those who are self-employed, working under-the-table, or employed outside of reporting jurisdictions may be missing.
Despite these limitations, the combined analysis offers a clearer picture than either data source alone — and a powerful starting point for asking deeper, more targeted questions about education, opportunity, and staying in Hawaiʻi.