Do the primary Mariel papers (Card 1990, Borjas 2015/2017, Peri & Yasenov) confirm the "no effect vs 10–30% dropout drop" figures?

claim-mariel-boatlift-wage-effect-flips-with-sample-construction rests on Silvia Merler's 2017 Bruegel blog post (Tier 2), a secondary retelling. The load-bearing numbers should be checked against the primary papers before the note can go beyond seedling.

What to verify, and where:

• Card (1990) — "The Impact of the Mariel Boatlift on the Miami Labor Market," Industrial and Labor Relations Review 43(2). Confirm the finding of no measurable effect on less-skilled Miami wages.
• Borjas (2015/2017) — "The Wage Impact of the Marielitos: A Reappraisal," ILRR (2017, NBER WP 21588, 2015). Confirm the ~10–30% wage drop and that it is specifically for the high-school-dropout subgroup.
• Peri & Yasenov — "The Labor Market Effects of a Refugee Wave: Synthetic Control Method Meets the Mariel Boatlift" (Journal of Human Resources, 2019; NBER WP 21801). Confirm the synthetic-control rebuttal and the "15–20 Miami observations per year" small-sample critique; check whether "very drastic" is Peri & Yasenov's phrasing or Merler's.

Why it matters: the whole hop rests on the claim that the sign of the result flips with sample construction. If the primary figures differ from Merler's retelling, the framing needs correction. Resolving this also lets claim-mariel-boatlift-wage-effect-flips-with-sample-construction shed its [unverified-quant] flag.