On December 4th 2021, Semeru Volcano in Lumajang, East Java erupted, emitting volcanic ash and lava. Fifty-four people died, 1027 houses were damaged and more than 9,000 people were evacuated.
Lumajang is now in a post-disaster recovery period, and Loop Indonesia visited Lumajang to conduct a survey using the Loop platform between 11th-14th August 2022.
Two Predikt staff and six enumerators from Taruna Siaga Bencana (TAGANA) under the Ministry of Social Affairs visited permanent houses around Sumbermujur Village and temporary shelters in Candipuro Districts over four days. Predikt was supporting the Indonesia Shelter Cluster during the emergency phase and the survey was conducted to find out what affected communities thought about the shelter assistance. In total, the Loop team gathered 399 stories.
The issues raised in the stories were quite consistent across women and men, and the requests for support and concerns raised were strongly linked. People had queries around when they would get new permanent housing and also what impact the location and type of housing would have on their livelihoods.
We learnt that housing without the opportunity for livelihoods was not positively impactful on the local population.
Over two thirds of respondents (212 out of 399) submitted a concern, whilst almost half (180 out of 399) submitted a request. Roughly 70% of all stories were about food security, while 39% of stories were about shelter. Of the stories about food security, 80% were about livelihoods, highlighting that this was a high priority and significant concern for the affected communities.
Some survivors now have to travel far to work, and some of them lost their livelihoods completely. A number of the community members complained that living expenses were now higher than before the eruption.
Children’s well-being was a concern, with 44 stories mentioning this; 28 stories noted that travelling to schools was too expensive, or that schools were not available.
Social aid was distributed, but was not considered to be delivered equitably. There are still community members who were unsure whether they would receive permanent housing or not, and only a minority managed to receive basic food aid.
What we learned from the process of using Loop:
Trained enumerators became more efficient with time, as they got used to the new approach. It was appreciated that they came from the local community.
Enumerators used primarily WhatsApp and a few used the weblink directly, based on personal choice.
Where possible enumerators showed local people how to use Loop on the beneficiaries’ own phones. For others, they preferred to use the enumerator's phone in the first instance. This helped to build confidence and knowledge of the platform.
Some stories were submitted in Javanese. This trend of people understanding the instructions in the main language but choosing to share their story in their own lesser-served language has been found in other contexts on Loop as well - Philippines, Zambia - and shows the importance of speaking in your own native language. Javanese was then translated into English. We hope to improve the platform to enable other languages when not solicited to also be available.
It took longer to moderate the stories due to Loop’s default filtering settings as well as the poor machine translation from Bahasa Indonesia to English. This required a lot of manual translation from the moderators.
In the initial days we saw the vast majority of feedback was coming from women because the men were out of the house - travelling for work etc. The team made adjustments to seek input in the evenings as well. The final result is a very equitable balance of male and female voices.
Most of the respondents were within the age of 30-59 years of age, but there were also a number of respondents over 60. This reflects the digital literacy of the affected population, as older people were less familiar with the chatbot function.
No people self-identified as having a disability. However, 13 of the stories were about protection issues affecting the elderly or people with health issues. In the future we will coordinate with organisations working with people with disabilities to ensure their stories can also be shared.
Two sensitive stories were reported about possible nepotism in the beneficiary lists, a sentiment that was echoed in several of the open stories stating that beneficiary lists were ‘unfair’. However there were very few concrete actions able to be taken due to the generalised and vague nature of the stories. A summary of the information (without identifying details) was shared with the relevant authorities with the consent of the authors.
Overall useful data resulted and has been fed back to the overall response through Predikt’s role in Information Management on the cluster system. Many systems-wide learnings need to be considered. Also, we continue to work to encourage organisations to understand the Loop platform and to respond to individuals directly, thus increasing the potential impact of reporting on Loop.
This has been a useful learning process to help with the use of Loop in other sudden onset crises which befall Indonesia, or other countries and to be integrated into other monitoring, evaluation and accountability systems nationally.