shared with Droplr
In order to tests these hypotheses, I use a series of paired t-tests to compare the differences between amounts allocated to similar applicants across the experimental conditions. The results for eleven t-tests are shown in table 2. Test 1 compares the first application in condition one with the same applicant in condition two, where only the applicant’s work history has been changed. On average, the applicant who had an “excellent” work history was awarded an additional $65, a difference that is significant at conventional levels. Test 2 shows that “laziness” is punished with a greater magnitude than hard work is rewarded, as applicant two loses over $170 when they are said to have a “poor” work history. Again, the test statistic leaves us with little doubt that the work ethic manipulation did not have its desired effect. Tests 3 and 4 look at the differences between white and black applicants solely on the basis of race. Neither test shows any significant difference meaning that white applicants are not rewarded any more than blacks on the basis of race alone. However, further tests reveal that racial prejudice amplifies rewards and punishments.

Tests 5 and 6 show that Emily and Keisha, when both are being compared to “lazy” Laurie, are rewarded very differently for their hard work. For her hard work, Emily sees an additional $123, a difference significant from zero at conventional levels. On the other hand, Keisha is rewarded with less than $10, a difference that is statistically indistinguishable from zero. What these tests (5 and 6) show is that Emily earns roughly ten times as much money for her hard work than Keisha does, all else equal. We can also test if the difference of these differences is significant. This is precisely what test 7 does, and it shows that such a large difference was unlikely to have occurred by chance alone. Now we have substantial evidence that hard work is rewarded in a manner that is tainted by race; but is “laziness” punished in a similar fashion?

Tests 8 and 9 repeat the process from above and show that while Keisha loses over $100 for being lazy, Emily loses less than $25. Again, the former difference is significant while the latter is not: only the applicant with the African-American name (Keisha) is punished for being lazy. The final test, test 10, presents another difference of differences test and shows that the magnitude of these differences is again unlikely to have occurred by random chance.7 While we can clearly see from table 2 that hard work is rewarded and laziness is punished, the interaction terms illustrate how race conditions deservingness. This is one answer to those who take the “principled politics,” or race-neutral values, perspective on welfare: blacks see no reward for their hard work, yet are punished for laziness at a disproportionate rate. As this experiment shows, the American norms of equal treatment cannot solely explain these results: Americans’ attitudes towards welfare are closely tied to modern racism.