A recently-hired employee’s median salary as a percentage of continuing employees’ median, each year. Below 100% means recent hires earn less than the typical incumbent (expected for faculty, where incumbents are more senior); a falling line hints at starting pay lagging.
The salary data reports funds paidin a year, not the annual contract rate, so a person’s first year is usually partial (they started mid-year) — verified here: the median first-to-second-year change is +35–71% but second-to-third is only +2–3%. So a “recent hire” is measured in their secondyear (first full year), and only when it directly follows their first appearance; that first appearance must be 2016 or later (2015 is the censored baseline). Everyone who appeared earlier is an incumbent. A first-appearance proxy still can’t see people who left and returned.