In 1979, prior to the Mariel flow, the black unemployment rate in Miami was 8.3 percent. This unemployment rate rose to 9.6 percent by 1981, after the Mariel flow. Before we conclude that the Marielitos were responsible for this 1.3 percentage point increase in black unemployment in Miami, however, we have to determine what was happening in comparable cities, cities that did not experience the Mariel flow. It turns out that black unemployment was rising even faster in the control group, from 10.3 to 12.6 percent (or an increase of 2.3 points)—probably because macroeconomic conditions were worsening during that period. If anything, therefore, it seems that the Mariel flow actually slowed down the rise in black unemployment, so that the difference-in-differences calculation (or 1.3–2.3) suggests that the Mariel flow was responsible for a 1.0 percentage point decline in the black unemployment rate.
The conclusion that even large and unexpected immigrant flows do not seem to adversely affect local labor market conditions seems to be confirmed by the experience of other countries. For instance, 900,000 persons of European origin returned to France within one year after the independence of Algeria in 1962, increasing France’s labor force by about 2 percent. Nevertheless, there is no evidence that this increase in labor supply had a sizable impact on the affected labor markets. 28 Similarly, when Portugal lost the African colonies of Mozambique and Angola in the mid-1970s, nearly 600,000 persons returned to Portugal, increasing Portugal’s population by almost 7 percent. The retornados did not seem to have a large impact on the Portuguese economy.
George Borjas (Labour Economics, 2016)
Despite the reputation Borjas has as an immigration skeptic, I was pretty impressed with his empirical review of the literature on the Mariel Boatlift AND the ongoing Min Wage Wars in his Labour Economics textbook. It was very balanced and objective; the quoted passage makes the man sound like Michael Clemens or Lant Pritchett (please don’t take the quoted passage as his final words on the subject - he goes on to discuss some of his reservations with the above results but I was still really impressed).
Directly after the quoted passage, there was also a great discussion on the contradictory nature of the labour demand curve implied by the Card min wage results vs the Card Mariel results: the curve cannot be simultaneously perfectly elastic and perfectly inelastic. His conclusion: take natural experiments with a grain of salt.