At the author's request, this corrigendum corrects the number of Rwanda and Burundi migrants present in Buganda in 1969 in Table 1 of the above publication.Footnote ¹ This table originally stated that 152,321 migrants from the former Belgian colonial territories of Ruanda-Urundi (‘R-U migrants’) were present in Buganda in 1969, with a male/female ratio of 2.5, and making up 6 per cent of Buganda's total African population. This number refers to nationals from Rwanda and Burundi enumerated in Buganda and is derived from the 1969 population census (as the notes and sources below Table 1 should have specified).Footnote ² While itself consistent with the source table in the census report, this number substantially underestimates the actual number of people of Rwanda and Rundi origin present in Buganda at the time. It is also not appropriate for comparison with the totals for earlier census years included in Table 1, for reasons that will be explained below. The actual number of migrants from Rwanda and Burundi and their children present in Buganda in 1969 cannot be ascertained with precision, but was likely in the range of 400,000 to 450,000. The author apologizes for the error made.

The amended estimate is based on the following steps. First, alongside nationality, the 1969 population census also reported the number of individuals enumerated in Buganda by ‘district or country of birth’.Footnote ³ In that year 271,752 Rwanda- and Burundi-born individuals were present in Buganda (significantly more than the number of ‘nationals’ from these countries), with a male/female ratio of 2.3, and corresponding to 10 per cent of Buganda's total population. Second, the total of 271,752 did not include children born in Buganda to migrant parents and is therefore still an underestimation relative to numbers from earlier (colonial) censuses, which were based on reported ‘tribal affiliation’ no matter the birth location (see the notes to Table 1). To make the numbers comparable across time, we need to account for children born after their parents arrived in Buganda from Rwanda and Burundi. According to the 1959 population census, ‘25–35 per cent’ of people of ‘Banyaruanda’ and ‘Rundi’ ethnicity were born in Buganda, and the remainder in Ruanda-Urundi.Footnote ⁴ We may expect that the share of Buganda-born children increased between 1959 and 1969, assuming that the preceding process of migrant settlement continued.Footnote ⁵ This warrants the use of the upper bound of the 1959 range (35 per cent) to derive an estimate of the number of individuals born in Buganda, but with parents born in Rwanda or Burundi. This yields an estimate of 146,328 individuals.

Adding up those born abroad and in Buganda, we arrive at an estimated total of 418,080, with a sex ratio of 2.3 (assuming a sex ratio of 1 for Ganda-born individuals) and corresponding to 16 per cent of Buganda's total population. Considering that some migrants adopted Ganda ethnic identities, that the size of the Rwandan refugee population in Buganda after 1959 is not precisely known, and that the number of Buganda-born migrants’ children in 1969 can only be indirectly inferred (as above), this number should be taken as an approximation.Footnote ⁶ Nevertheless, it is more soundly derived and therefore more meaningful than the number originally used in Table 1, which was far too low.

The repercussions of this correction are as follows. Rather than declining, as was previously suggested in the article, the size of the Rwanda and Burundi migrant community in Buganda stabilized between 1959 and 1969. The corrected, higher number no longer warrants the inference that ‘many of the families who had earlier settled in Buganda had left’.Footnote ⁷ Otherwise, the use of this more appropriate number does not affect the arguments pursued in the article. The growth of the Rwanda and Rundi migrant community in Buganda clearly tapered off after 1959 and was very modest relative to the increases recorded by earlier censuses, as well as relative to the growth of Buganda's overall population between 1959 and 1969. This is consistent with the arguments pursued on pages 401 to 405.