Food risks in PNG: lessons from 1997

Sweet potatoes for sale, Intoap market, Markham Valley, PNG in 2010 (image: Flickr/Biodiversity International/P Mathur)

The current El Niño event in the Pacific Ocean is disrupting food production in Papua New Guinea and Indonesian Papua. Whether the current El Niño event will continue to develop for the rest of 2015 is uncertain, but likely. Nor is there a sure way of predicting the severity of food shortages in PNG from the severity of an El Niño event. But already this year the impacts on food production in PNG have a worrying similarity to the way events unfolded in 1997, when food and water shortages were more serious than they had been since 1941 and possibly since 1914.

How does El Niño affect PNG food production?

Around 80 per cent of food consumed in PNG is grown there. A strong El Niño event reduces rainfall significantly in normally wet areas for some months and prolongs dry seasons. This is because PNG’s characteristically high rainfall is the outcome of low air pressure, upward movement of air in the Walker circulation, onshore winds crossing warm seas and picking up large amounts of moisture, and high mountain ranges forcing the warm moist air up in altitude, where it cools and the moisture condenses and falls as rain. During a strong El Niño event, the western Pacific cools and the Walker circulation reverses so that cool, dry, high altitude air descends over PNG. The result is less than average rainfall for some months and significantly less cloud cover. At altitudes above 2,200m the lack of cloud cover at night allows temperatures to fall to below zero and frosts to occur.

Historical records reveal some of the impacts of El Niño. There were 11 droughts associated with El Niño events in PNG between 1896 and 1997, but severe disruptions to food and water were reported only in 1902, 1914, 1941 and 1997. In 1972, the Australian colonial administration, in ignorance of how high altitude people move to lower valleys in response to frosts, decreed people should not move and the Australian Defence Force delivered food by helicopters. Following 1972, stories of a more severe highlands food shortage in 1941 began to be heard. A post-1972 examination of newspapers and administration reports found a probably even more severe event occurred in 1914. But the pre-1972 documentary evidence was fragmentary and excluded the highlands, where there was no colonial administration until after 1945.

The 1997 event was probably the most severe since 1914. At altitudes above 1700m, repeated frosts killed all sweet potato, the staple food. At lower altitudes, newly planted gardens failed because of a lack of rain and because streams used to make sago dried up. Taro growers were forced to remove plants from gardens and place them near large streams to maintain planting stock. Wildfires destroyed villages and killed people and pigs. Large areas of montane forest trees killed by fire in 1997 are still evident on highlands mountain ranges today. Forest fires disrupted long fallow shifting cultivation systems between the Fly River and the Sepik River (D. Jorgensen, pers. comm.) where rainfall averages as much as 8,000 mm per year. The Fly River became too shallow for shipping and Rouna hydro-power station was closed to preserve Port Moresby’s water supply.

By December 1997, national surveys of food and water found that over 260,000 people were eating ‘famine’ foods such as wild yams, tree leaves and banana roots. A further 980,000 were eating small amounts of poor quality garden food. Rural people survived by employing a number of strategies: eating ‘famine’ food; purchasing imported food with cash raised by selling off pigs or from savings; moving to places where food was available; moving to towns to stay with relatives who had wage-paying employment; and receiving rice, flour or cash from relatives in employment. Australia used ADF aircraft to reach 100,000 people without food, living in places only accessible by air. But, as a result, the numerous company and mission light aircraft that regularly flew to the same airstrips used by the ADF were grounded and a number were taken to Indonesian Papua to provide relief transport there.

In 1997 rainfall deficits were greatest furthest away from the Equator, but the impacts of the food shortages on people were greatest in the poorest parts of the country. People with wage employment, with relatives in employment or with adequate savings cared for themselves by purchasing imported food. The best evidence for this is rice import figures that show an increase of 66,000 tonnes in 1997. Of this additional rice, 76 per cent was sold through retail outlets. The balance was purchased by the Japanese, Australian and PNG governments and was delivered as relief food [pdf] (p. 155–63). Most people in PNG had the impression that Australia had rescued them from a catastrophe. In fact, they had largely saved themselves [pdf].

What is happening in PNG now?

The first signs that something was amiss in past PNG food shortages caused by El Niño were press reports of frosts. This year social media postings preceded press reports. Facebook postings, in particular, contain photographs of frosted gardens at Tambul and Kandep. Mobile phones are now widespread in rural areas. The National Agriculture Research Institute’s (NARI) Tambul station reported frosts on the nights of 19 and 26 July and 10, 11, 12, 13 and 14 August, a very similar pattern to 1997. In 1997 frosts continued into September and early October.

The 2015 frosts have had an immediate and spectacular impact on sweet potato and potato crops. Across whole valleys, fields of plants have died and gone brown although cabbages, peas, brassicas and wheat have survived. As they did in 1997, media reports describe deaths of infants and the aged killed by frost and an undefined general illness caused by eating food which has been exposed to the frost. These reports ignore that fact that deaths of infants and old people occur all the time in rural PNG and eating sweet potato harvested shortly after a frost does not cause sickness. Nevertheless, people face serious problems because a number of frosts some weeks apart will stop production for up to six months from what is a continuous planting and harvesting system.

For the next six months life will be tough in the frost-affected communities, but they will not passively wait for help to come to them. Some are reported to be already on the move to lower altitude valleys, where they maintain strong social relations, especially through marriage, as an insurance policy against frost. When the frosts cease and rainfall is again received, they will go back to replant their sweet potato gardens with vines from the lower valleys. Survival is not guaranteed, however. In 1997, widespread drought also severely reduced sweet potato production in the lower valleys to the extent that the additional refugee human and pig populations could not be adequately fed. If they had relatives in towns who were earning cash, many people moved into the towns, or received cash and food sent by urban relatives. The food supply situation in the lower valleys is not known at this time.

In August, media reports began to appear of low water levels in mainly highlands rivers, as well as the Fly and Sepik rivers. The Ok Tedi mine responded to the low Fly River by closing operations, a decision that was also related to low copper prices and a mine collapse. The drying of smaller streams is a particular problem for people who use sago as a staple because an ample supply of water is needed for leaching starch from the palm.

The national newspapers are reporting below average rainfall from many places, but there are no reports of widespread food shortages from lowlands villages yet. It needs to be borne in mind though that the communities which are least able to get themselves through a food shortage are isolated from roads and urban markets, and so cannot earn cash and buy imported food like rice to carry them through. Their very isolation also means their situation is less likely to be reported. A number of reports have been published of schools and health centres about to close because of lack of water for teachers and pupils. In 1997, the majority of schools had water tanks that were rusted out and empty. Teachers collected water from small streams nearby until they dried up, forcing them to leave. Recommendations that the tanks be replaced before the next drought have been ignored.

Government responses so far have been largely uncoordinated. At Kandep, the Leader of the Opposition in the national parliament, Don Polye, set up a committee chaired by an unnamed priest and gave it K2 million to buy rice, flour and other food. Enga and Simbu provinces have set up “trust accounts” and invited contributions. The Southern Highlands Governor has appealed for help from the national government, while the Enga Governor has urged Engans “not to panic”, because help is on the way. On 19 August the national government convened a Government Action Group in Port Moresby after which the Chief Secretary to Government, Sir Manasupe Zurenuoc, promised “immediate funding” on receipt of assessment reports from districts and provinces. NARI has been tasked to send a team into the frost-affected areas to make assessments. The Director of the National Disaster Centre (NDC) announced that NDC had no funds to deal with the situation, but that if it should receive some they would be properly accounted for.

What is to be done?

The present situation in the highlands needs urgent attention in the form of a proper assessment of food availability. A person with qualifications in agronomy and experience of the 1997 event should be appointed as a national drought coordinator. Teams of agriculturalists with experience of highlands food production systems and the impact of frosts on sweet potato should carry out assessments. A source of funds to provide aid, if it is required, should be identified and secured. Steps should be taken to prevent unjustified price rises of imported rice and wheat. As much as possible food, rather than cash, should be provided to MPs or public servants for distribution. The actual distribution of food should be overseen by the coordinator because in 1997 a lot of food was dumped on airstrips or small rural centres by MPs and people were left to walk long distances to collect it. Many old people were unable to make the journey to the distribution points. Lower level airlines and mission aviation can be used to distribute assistance to isolated places, rather than military helicopters. People should not be prevented from moving, and old people and very young children could be assisted to move. The Government of PNG should liaise with its regional partners such as Australia earlier rather than later, and regional partners should maintain close watching briefs in case weather conditions worsen, frosts continue to occur and rainfall is not received in December.

SSGM logoBryant Allen is a Visiting Fellow in the State, Society & Governance in Melanesia (SSGM) Program at ANU. He wishes to acknowledge comments from Michael Bourke. This is an updated and expanded version of SSGM In Brief 2015/39, published on 14 August 2015.

Bang, S. K., S. Poloma and B. Allen 2003. Stabilization of Upland Agriculture Under El Niño-Induced Climatic Risk: Impact Assessment and Mitigation Measures in Papua New Guinea. CGPRT Centre Working Paper No. 73. Bogor, Indonesia.

Cobon, D. H., M. Ewai, K. Inape and R. M. Bourke 2015. Food shortages and extreme climate events in Papua New Guinea. (unpublished).

image_pdfDownload PDF

Bryant Allen

Bryant Allen is an honorary associate professor in the Department of Pacific Affairs at The Australian National University.


  • Thank you Bryant for this important post. It is also positive to see Vini is drawing on the considerable material that fortunately was documented from the 1997 experience. When going through the historic material, it is not clear what level of public resources were applied to help mitigate the impacts of the drought and frosts. The 1998 budget papers suggest an allocation of K25m for drought relief, K38m for additional education subsidies (aimed to lessen the impacts of the drought on local cash flows) and K8m for medical supplies. Donor expenditure is unclear – one figure is AUD30m in relief from Australia. Other donors were involved, as were NGOs, church and community groups. Given inflation from 1997 to 2014 of 240% (using CPI indexes from the IMF), a population increase is some areas of 50%, and an AUD to Kina exchange rate of around 2 to 1, the total budget requirement may need to total K668m ((K25m+K38m+K8m+K60m (donors)) x 3.4 (inflation) x 1.5 (population)). This is a rough calculation, but it clearly suggests the current allocation of K30m is inadequate – less than one-twentieth of an indicative value of the 1997 response. However, given the fiscal pressures the government is now under due to spending LNG revenues before they were received, and then the collapse of international commodity prices, there appears little fiscal space to move. However, the funds were found in 1997 despite fiscal stringency in place under an economic reform instigated in 1995 due to poor economic management. Hopefully, public funds can be better spent this time through using private sector supply chains, traditional coping mechanisms can be drawn upon, and more emphasis can be given to actual need rather than politics. Agriculture prices also moved upwards during the 1997 crisis which more than offset falls in quantities driven by the drought – this helped support local incomes (especially coffee but also copra and cocoa – although oil prices fell to US14 per barrel). It is unclear whether this beneficial offset through agriculture price rises will occur in the current drought/frost disaster. One way to accomplish a similar effect is to allow the Kina to depreciate more quickly (which will increase agricultural incomes as exports become more valuable in Kina terms). Given the potential immediate humanitarian impacts, the medium-term impacts on the agriculture sector, and the possible long-term impacts on children’s health if inadequate nutrition is supplied during their growing years, there is a need to urgently assess the required level of public resources needed and the macroeconomic levers which may be able to assist. In addition to domestic action, this may include having to go to the international community to request assistance.

  • Hi Bryant, Thank you for bringing up this discussion on the ANU Policy Blog. I do agree with you that the impact of El Nino on food production in PNG is quite worrying. I also like the analysis that you have provided on food and water shortages around PNG including the history of similar events over many decades. Compared to the response of 1997-98 drought and frost, the Government of PNG is more aware of its roles and responsibilities in this situation this time around. The print and social media have done well in highlighting the impacts to the rest of PNG and the authorities. Using this information together with assessments by various provincial authorities, the PNG National Disaster Centre was able to target its ground- based assessments. Thanks to the work of people like yourself and Mike Bourke, we are learning from the lessons of 1997-98 and putting in place better strategies to address the impacts of El Nino on the local communities. We do acknowledge that there may be gaps in the current response but at least we are more aware of what to expect and how as a responsible Government can address the situation. We have already reached out to the development partners including the Government of Australia for specific assistance but the biggest part of the response will be handled by Papua New Guinea government sectors this time around.

    To give a bit more insight to the Government’s response, we have already deployed an assessment team to the highlands region while the next two teams to the Momase and New Guinea Islands regions will be deployed by end of this week. Those teams comprise of trained people in health, agriculture, protection, nutrition education, logistics and WASH. NARI is represented in all the teams. So a rapid assessment of the situation is currently taking place. This is accompanied by initial relief assistance especially to areas affected by frost. Based on the assessment reports, we will be able to have a more targeted response. Detail sector assessments will be carried over the course of the El Nino period to allow the Government to monitor the situation and response accordingly.

    We are confident of our efforts this time but acknowledge that we still need the support of our bilateral and development partners. Every El Nino event has its own lessons and we are grateful for those of the 1997-98 event that people like yourself have documented well. In fact, I am using your publications as reference materials in supporting the Government’s efforts.

    Vini Talai, PNG National Disaster Centre

    • Vini, Many thanks for that detailed response. It’s hard to learn what is happening from the newspapers reports. Good to see you are working with NDC. Best wishes, Bryant

      • please am very pleased that the current government is taking the initiatives to address the current issues that are experience in the highlands of PNG and am very concerned about my very dry place IALIBU ..thank you very much.

Leave a Comment