"Estimating The Aboriginal Population In Early Colonial Australia: The Role Of Chickenpox Reconsidered' by Boyd H. Hunter and John Carmody in (2015) 55(2)
Australian Economic History Review 112–138 comments
Noel Butlin radically altered the debate about the pre-colonial Aboriginal population when he provided a set of hypothetical demographic scenarios, which nonetheless were both grounded in economic theory or human ecological considerations and broadly consistent with what we know about the historical record. This research builds on Butlin’s legacy by exploring how his scenarios are consistent with both the medical understandings of the infectiousness and mortality of various diseases and the history of settlement. Another contribution from this paper is to highlight the possible role of chickenpox in the Aboriginal depopulation in the early colonial period.
The authors state
After a prolonged period of coastal exploration of Australia by Europeans and Asians, the substantial process of colonisation was based on a unilateral appropriation of Aboriginal land by Britain in 1770. Though no war was ever declared, the frontier violence started shortly after arrival of the First Fleet and persisted well into the twentieth century. Like any war, declared or otherwise, the conflict led to many deaths on both sides. Nevertheless, the number of Aborigines who were directly killed as a result of such violence is likely to have been dwarfed by the spread of introduced diseases such as smallpox and respiratory diseases.
Many researchers have attempted to estimate the pre-colonial population. One of the first credible systematic estimates of the Aboriginal population was provided by Radcliffe Brown who argued for ‘the original population of Australia having been certainly over 250,000, and quite possibly, or even probably, over 300,000’; furthermore, he deliberately provided conservative estimates because the ‘data are scanty and for the most part unreliable’.
Noel Butlin provided a fundamental reassessment of the size and nature of the Aboriginal economy and population before 1788, with a particular focus on demographic modelling that was informed by some of the key economic features of the Aboriginal economy. Butlin extrapolated numerous detailed demographic scenarios based on extensive detailed knowledge of the early colonial history in the south-east of the continent. He convincingly argued that the early estimates were likely to be substantial under-estimates, and he eventually concluded that the pre-colonial population was between 1 and 1.5 million. Some of Australia’s leading archaeologists, notably John Mulvaney and Peter White, have argued that Butlin’s extrapolation of the circumstances and conditions in the colony of New South Wales was not warranted; nonetheless, they used some of his parameters to argue that the pre-colonial population was more likely to have been between 750,000 and 800,000 Aborigines. Over time these estimates became known as the ‘Mulvaney consensus’ about the best available estimates of Australia’s population before 1788. While most researchers have given considerable credence to such estimates, it is not entirely clear which set of Butlin’s parameters had been used to establish the consensus, except that the assumed rate of depopulation was conservative and that the mortality rates from diseases were relatively low outside the areas that were extensively settled by European colonists before 1850.
Butlin’s analysis was rigorous, yet there is clearly a wide variation in the possible Aboriginal population implied by demographic parameters provided in his research. Furthermore, there is some fundamental and unavoidable uncertainty about the basis of the calculations. Butlin’s Economics of the Dreamtime was subtitled a ‘hypothetical history’ and it is a legitimate exercise to examine the implications of the various assumptions outlined in that analysis. This article argues that it is necessary to specify clearly all of those underlying assumptions to enhance the credibility of the estimates. Particular attention is paid in the following analysis to the transmission of disease and associated mortality rates, but the article also reflects on the role of resource depletion associated with the dispossession that accompanied what was in some ways a military invasion.
This article argues that one disease is usually ignored in the existing calculations of the pre-colonial population, chickenpox. While smallpox probably killed many Aborigines, chickenpox is much more infectious than smallpox and therefore could potentially explain the apparent ease of transmission of the diseases across the less densely populated regions of the Australian continent. Less infectious diseases also have a role to play, especially if they have a high mortality rate. Some diseases can only be picked up by extensive contact with infected people, and hence the effect of such diseases depends on the proximity and exposure to infected people as well as the infectiousness of the disease and mortality rates for those who acquire the disease. That is, the historical progression of relatively dense settlements throughout the continent needs to be taken into account for less infectious diseases with high mortality rates to play a significant role.
The next section documents some of the range of demographic parameters and the possible population trajectories implied by those parameters. It also revisits some salient history that could inform the scenario-building exercise. The following sections explore problematic assumptions for estimating the pre-contact Aboriginal population with a view to motivating why we need to re-estimate the original Australian population. The plausibility of estimates hinges on the extent and timing of the transmission of disease, and the mortality of those diseases in the Indigenous communities at risk. The tension between various assumptions provides a reason to explore the role of chickenpox, which was not considered in Butlin’s simulations. After we provide an argument why that disease should be considered, Butlin’s scenarios are revisited with explicit assumptions about the transmission of disease partially informed by the history of settlement. The last two sections of the article examine the potential insights for the economic history of immediate post-contact period.
By imposing a greater consistency on the assumptions about transmission, infectiousness and mortality of diseases, we will eliminate less plausible hypotheses. Even after implausible hypotheses are eliminated, the pre-contact population can take on range of values, and it is a matter for debate and further research to establish a robust consensus about the most reliable population estimates. One of the important findings of this article is that the chickenpox hypothesis is both consistent with plausible assumptions and provides a relatively straightforward explanation for the ease of transmission of disease that may have killed so many Aborigines in the early days of European settlement.
They conclude
This paper argues that there is a need for greater transparency and consistency in analyses of the economic history of Indigenous Australia. Most of the existing research on the pre-contact population provides only rather sketchy detail of how estimates are derived. Our analysis demonstrates that these details matter, in that the pre-contact population estimates can vary from 500,000 to 1.2 million. However, by being transparent in the assumptions required to construct the population, it should become apparent to the reader that some estimates are more plausible than others.
It is important that the preferred population estimates are consistent with both the known historical facts and current medical knowledge. Chickenpox is introduced in this paper as one of the possible main causes of Aboriginal mortality in this period because it is highly infectious and hence is plausibly a disease that could facilitate fast transmission of serious illness across the entire continent. The chickenpox scenario is also consistent with the ‘Mulvaney consensus’ about the pre-contact population.
The ‘History Wars’ have made it difficult to talk about the extent of frontier violence and warfare in Australia. The inadequacy of official records and existing evidence means that the issue will continue to be contentious. Notwithstanding, it is important to acknowledge related issues that are potentially significant for historical estimates of the Aboriginal population. This paper uses Butlin’s estimates of the effect of ‘resource loss’ as a proxy for what was, in essence, an (economic) conflict over scarce resources. The primary resources lost were both productive members of, and the land used to sustain, the local tribes. The potential population effect of ‘resource loss’ during the colonial expansion appears to be substantial. Clearly, ‘resource loss’ needs to be taken into account in any estimate of the Australian population in this period.
The estimates provided in this paper can vary substantially from the current consensus of the pre-contact Aboriginal population. The estimates based on the Campbell’s, Invisible Invaders, seem to lead to excessively high values for the population across Australia. Our preferred estimate is exactly equal to the Mulvaney estimate of 800,000 people.
The medieval Franciscan friar, William of Ockham, outlined the principle that the hypothesis with fewest assumptions should be preferred over more complex formulations. In our opinion, the hypothesis that chickenpox was responsible for a substantial numbers of Aboriginal deaths in the early colonial period provides the most straightforward explanation. More importantly, it also has a solid scientific basis in epidemiology. Not only does it provide assumptions that are consistent with what we know of the transmission, infectiousness and mortality of diseases, it does not rely on conspiracy theories. While Ockham’s razor can be used to support the proposition that chickenpox played an important role in the early European settlement, it is clear that other diseases played a role in the later colonial period. Whichever diseases were involved, it is important to ensure that assumptions about the epidemiology and medical knowledge about those diseases are accurate and the population estimates are consistent with the historical record. Ultimately, the research in this paper enhances our confidence in the Mulvaney consensus by allowing us to eliminate implausible scenarios. Of course the precise explanation for the population loss in this period is still open to debate.
