Which of the following are conclusions from Moro's paper on statistical discrimination?

• Statistical discrimination can lead to both efficient and inefficient outcomes.
• If firms statistically discriminate, then over time they'll learn more about the workers' productivity and rely less on the group mean productivity to determine their wages
• Groups that are discriminated against could achieve higher productivity if employers didn't treat them differently than the privileged group.
• If firms statistically discriminate against Black workers because they think they're not as productive as White workers, Black workers have a lower incentive to invest in the skills required to become more productive, and so they are likely to become less productive than White workers even if they hadn't been previously.

https://andreamoro.net/assets/papers/palgrave-statistical-discrimination.pdf