Counting the 1918 Flu Dead in a Small Town
Whakapirau, a North Island town of about 1,400 people, is preparing a centenary panel on the 1918 influenza pandemic. The official death register lists 43 deaths: 38 Pākehā and 5 Māori. The local newspaper of November 1918 reported 'over 60 deaths' and named several Māori families whose deaths do not appear in the register. Two kaumātua oral accounts, recorded in 1995, say that 'many more than five' Māori died and that some families avoided official notice because they feared removal of bodies and because trust in authorities was low after land confiscations. The public health officer's returns for six weeks are missing; a note says the officer was ill. The panel is 200 words, with no footnotes, and the council wants 'a clear number' for visitors. The historian has four weeks and no budget for further archival research.
1. Explain why the official register and the Māori oral accounts diverge, and what this divergence indicates about the evidential status of each source.
Read the worked answer
The divergence is not simply a case of one source being 'biased' and the other 'true'. The death register was a state instrument: it recorded deaths that were reported to officials, usually through a doctor, undertaker or family member. In 1918 the epidemic overwhelmed local officials, and the missing six weeks of returns show that the administrative machinery itself broke down. For Māori, under-registration is also historically plausible because hospitals and official burial procedures were viewed with suspicion after land confiscations and because some families deliberately avoided state notice. The register is therefore strong evidence of recorded deaths, not of total deaths.
The oral accounts serve a different evidential purpose. They are not statistics, but testimony about experience, community memory and official neglect. Their value lies in explaining why the register is incomplete, not in supplying a precise alternative count. A defensible reading treats the divergence as itself historical evidence: it reveals unequal access to state recording and unequal trust in the state. The historian should neither dismiss the register as a lie nor treat oral testimony as a more authentic count. Instead, each source must be read for what it was designed to do and for the silences it produces.
2. Recommend how the panel should present the death toll. Justify your recommendation with reference to the sources and the panel's constraints.
Read the worked answer
The panel should avoid a single 'clear number' because no source supports one. A defensible formulation would be: 'At least 43 deaths were officially recorded in Whakapirau, including 5 Māori. Contemporary reports and Māori oral accounts indicate that the true toll was higher, especially among Māori, whose deaths were less likely to be registered.' This wording does three things. First, it states what the register can prove: a minimum figure. Second, it signals that the minimum is not the total. Third, it names the groups most affected by under-recording without pretending to replace missing data with a false precision.
The 200-word limit is not a reason to omit the problem; it is a reason to compress it into one carefully chosen sentence. The historian can use the phrase 'recorded deaths' rather than 'deaths', and can explain in a short clause that the public health returns are incomplete. This is more ethical than a confident round number, because it refuses to make the marginalised disappear twice: once in the records and again in the exhibition. It is also more historically defensible, since it makes the evidential limits part of the public interpretation rather than hiding them.