And it may be out for some time yet. The process of self-recovery is shifting and elusive, little understood, and difficult to assess, occasioning imperfect decision-making. The kind of conclusions that are reached in the rush of an emergency response, when proposals are elaborated and costed within days, are inevitably imprecise.
Rebuilding begins immediately
Cyclone Pam caused extensive damage to housing in Vanuatu. Typically, housing is made of poles, bamboo, wild cane and palm thatch. In many villages 80 to 90% of these homes were destroyed by the storm. Within days, houses were being rebuilt and it was immediately evident that self-recovery was going to be fast-paced and generated by the immediate needs of the families. Initial assessments showed immense damage, but also very significant recovery.
This presented an immediate challenge to the humanitarian community: how best to respond to the evident immediate need for emergency shelter; but also how to support the early recovery, self-build process that was happening in parallel.
How accurate was this interpretation? Can self-recovery be understood so quickly and easily? The first field visits, village walks, and community conversations all concluded that people were rebuilding their homes. The much more rigorous REACH assessment conducted a few weeks later came to similar conclusions, reinforcing the first impressions.
Without a doubt construction was happening very fast as the families sheltering in schools and churches returned to their villages. We responded as quickly as possible with tools, nails and cyclone strapping to support this process, coupled with training in simple safer building techniques. However, three months after the cyclone, the pace of rebuilding had slowed down considerably and we were still distributing the same package of tools and materials.
The dynamics of self-recovery
There are various theories about what is happening in the self-recovery process. Much of this is anecdotal; none is based on rigorous evidence. Almost certainly no generalisations can be drawn, with a number of different personal and contextual circumstances coming into play. Nevertheless, some factors might include:
- The original flurry of self-building was never seen as being permanent rebuilding, but rather rapid but temporary shelter for the time being.
- Good roofing material was in short supply because of the storm and the re-use of salvaged roofing material, or sub-standard palm thatch, was seen as make-shift rather than permanent.
- With immediate shelter needs met, agriculture became the priority; after a short while the pace of construction slowed.
- A further bout of construction might occur as the cold, wet months of July and August pass and the cyclone season approaches.
- The fixing kits – nails, cyclone strapping and so on – are the only materials that would have to be purchased and are considered high-value. They are not going to be used on a temporary house, but kept until the family makes a permanent dwelling.
- A temporary house might be intended to last six months, but might still be there in three years’ time; does this constitute self-recovery?
The need for constant (re)assessment
None of this shows that the original assessment was wrong: self-recovery was happening fast. However it does suggest that the analysis and interpretation needs to be constantly re-evaluated and that the priorities of the population may change as time goes by.
Initially it seemed urgent to get good building messages out to the villages before a significant amount of building had been completed. However, perhaps as the pace of building slowed this became less important. Indeed it could make more sense to delay, prepare better, and roll-out good safer building training in time for the next round of house building. (This might also resonate with the Nepal response where the monsoon slowed down recovery and allowed time for a good recovery strategy to be put in place.)
Can we do smart programming that takes these uncertainties into account? Can we be nimble enough to tweak and alter programming to respond to deeper understanding and changing circumstances?
Ingenuity and recovery
In Vanuatu, the original distribution of tarpaulins proved to be a valuable self-recovery element as well as an essential emergency response. They had a variety of uses: patching up roofs due to a shortage of palm and natangora; enclosing a latrine; walling a kitchen. This is testimony to the versatility of tarpaulins: as a relief item, they certainly allow for families to use them in different ways according to their changing needs.
More by force of circumstance than planning, the second wave of training did not happen during the emergency shelter response. This was to be the training of shelter ‘ambassadors’ or champions, who would then be charged with promoting good, safe building practice in the villages. Now this aspect of the programme has been pushed into a recovery phase and will eventually be incorporated into the long-term resilience work of CARE Vanuatu. It is possible that this is all for the good. Perhaps careful assessment now would determine if and when the families are likely to rebuild or improve their houses. This might give us an indication of when the next round of training should occur.
This is written after the event and from the other side of the globe. Much is conjecture. However, Vanuatu does illustrate the need for constant (re)assessment, keeping an ear to the ground, being in conversation with the community, listening to partners and being always ready to seek new directions.