It is estimated that between eight and nine million people live in Papua New Guinea, but no one knows for sure. Honorary Associate Professors Mike Bourke and Bryant Allen from the Australian National University explain why PNG’s last census was a failure and what can be done to avoid another potential flop this year.

Goroka, Papua New Guinea. Credit: 9eGuide Travel/Flickr

What is the population of PNG? The short answer is nobody knows and there is no official estimate of the population of the country in 2020.

Estimates vary among sources but are usually between 8 and 9 million. For example, a figure of 8 million is used in the PNG Agriculture Medium Term Development Plan 2020–2022 developed by the National Department of Agriculture. A paper in early 2020 by the National Research Institute projected provincial populations totalling 8.943 million for 2020, with a caution that their estimates should be treated as minimum figures. That figure is similar to that given by the United Nations of 8.947 million, presumably because similar techniques were used to generate those estimates. The CIA World Factbook gives the population as 7.3 million in July 2020.

The censuses conducted in 1980 and in 2000 are generally considered to provide reasonably good estimates of the provincial and national populations. There has not been a reliable national census for over 20 years; the 2011 national census was considered a failure by the PNG government.

Knowledge of the national population of PNG is important. It is needed to calculate statistics required for planning purposes. These include the death rate, population density, food consumption, employment and income. It is also necessary for the provision of services including schools and health facilities.

‘Until there is a national census that produces reliable data, we will not know the national population.’

PNG national censuses have occurred in 1966, 1971, 1980, 1990, 2000 and 2011. The population has grown from 2.2 million in 1966 to 7.3 million by 2011. The population is growing rapidly, but the reported growth rates vary for different inter-census periods and are in the range of 2.2 to 3.1 per cent per year.

In a recent Devpolicy Discussion Paper, we generated projections of the mid-2020 population. We used the base population in 2000 (5.19 million) and three growth rates to create a population range. We concluded that the national population in mid-2020 was likely to be between 8.8 and 9.6 million people. Experience in the 2015–2016 drought was that the national census data sometimes underestimated the actual population. Hence, we suspect that the actual population in mid-2020 might be closer to the top of that range, that is about 9.5 million people, but until there is a national census that produces reliable data, we will not know the national population.

Urban populations

Australian National University’s Bryant Allen. Credit: Devpolicy

The proportion of people living in urban areas in PNG is one of the lowest, if not the lowest, in the world. This has been consistent in each national census, with 12–13 per cent of the total population living in designated urban areas.

Most PNG towns have spread physically beyond their urban census unit boundaries into surrounding rural census units. The outcome is that enumerated urban populations are probably lower than the real numbers of people who live in towns in PNG. This may explain to some extent why the proportion of the population who live in towns is not reported to be increasing. Most urban boundaries need urgent review before the next census, but this would be a major exercise.

There are two larger cities in PNG (Port Moresby and Lae), eight moderately large towns with a population between 20,000 and 35,000 people in the 2011 census (Madang, Mt Hagen, Kokopo-Vunamami, Popondetta, Wewak, Goroka, Kimbe and Mendi). As well, there are another 17 urban centres with a population between 4,000 and 17,000 people in 2011. (Note that the reported population of Tari of 39,000 is almost certainly an error.) There are about a further 50 small to very small centres classed as urban in PNG. However, the total population in those centres is small with only about 60,000 people living in urban centres with a population of less than 4,000 people in 2011.

‘Given the likely underestimation of the population in many urban centres because of out-of-date urban/rural census division boundaries, it is possible that the urban population could be as high as 1.5 million.’

The urban citizen population in the 2011 census was 902,900 people, with a further 20,900 non-citizens living in PNG, many of whom lived in urban locations. The urban population in 2020 is not known but would be well over one million people. Given the likely underestimation of the population in many urban centres because of out-of-date urban/rural census division boundaries, it is possible that the urban population could be as high as 1.5 million. Again, we do not know this figure.

As well as those living in designated urban centres, about six per cent of the national population live in ‘rural non-village’ situations. These are rural schools, mission stations, plantations, oil/gas and mining camps, logging camps and agricultural research stations. These are in rural areas, but they are not villages, hence the term ‘rural non-village’. 

The next census

Australian National University’s Mike Bourke. Credit: Devpolicy

The national census was scheduled for 2020 but was deferred to 2021 because of the COVID-19 pandemic. Only limited preparations have taken place for a census in 2021. Whenever the next census is held, it is most important that it is conducted carefully and competently. PNG should seek assistance from international experts in censusing countries in similar conditions. It will require international funding to provide the resources needed to plan in advance, do the counting and analyse the results.

It is worth considering whether the national census should occur over the whole country within a few months, which spreads the expertise thinly across the whole nation and is difficult to properly supervise. Collecting census data over 12 months, with fewer well-trained and well-resourced teams moving progressively from province to province, working closely with district staff, ward councillors and locally employed field staff, could deliver a completer and more credible census.

Read the full discussion paper at devpolicy.org

This article appeared first on Devpolicy Blog, from the Development Policy Centre at The Australian National University. R.M (Mike) Bourke is an Honorary Associate Professor in the College of Asia and the Pacific at the Australian National University (ANU). Bryant Allen is an Honorary Associate Professor in the Department of Pacific Affairs at ANU.