December 13, 2011
An independent evaluation of shelter provision released last week by Estudios Proyectos y Planificación S.A., under commission of the International Federation of the Red Cross, provides perhaps the first systematic evaluation of the provision of shelter since the earthquake nearly two years ago. The report, while acknowledging the tremendous constraints in post-earthquake Haiti and pointing to some notable successes, is highly critical of the overall effort on the part of the international community despite the fact that “money was not an issue for the shelter response.”
The report focuses on the Shelter Cluster, which took the lead in providing emergency and then interim shelter solutions in Haiti, finding that affected populations and Haitian institutions were excluded from the process and a rigid, singular focus on transitional shelters (T-shelters) hindered the ability to develop a comprehensive housing solution.
Meetings were most often conducted in English and access was restricted inside the UN Log base leading to “a barrier between the international response system and the Haitian institutions.” One government official states that, “[o]ur ideas were not taken too much into consideration. Some said it is because we didn’t have the capacity [to actively participate in the cluster’s decisions] (…) Perhaps we were weak but we were there and tried, but they [shelter agencies] wouldn’t listen to us.”
The evaluation found that “a more participatory strategy would have been desirable to better address the affected population’s needs and plans and to seek collaboration with them, to allow a more self-driven response and to reduce the burden on the humanitarian actors.”
“Affected people were not consulted nor their capacities considered, the response was what those with the [foreign] money decided,” one interviewee told the evaluation team.
The Shelter Cluster strategy, developed soon after the earthquake, focused on two stages: an emergency phase lasting three months and then an interim phase which aimed to provide full coverage of transitional shelter within 12 months.
The evaluation notes the successful distribution of tarpaulins in the first four months; however, because the interim strategy was significantly delayed, the “shelter sector did not accurately measure follow up emergency shelter needs” or integrate “reinforcement/replacement actions” into a comprehensive strategy.
Additionally, while the Shelter Cluster reported shelter coverage of over 100 percent, HRRW noted at the time that this did not take into account gaps in coverage at certain locations, resulting in an estimated 232,130 people being without any sort of shelter five months after the earthquake.
The resulting conditions in the camps as the rainy season neared were untenable. It was found that “the protection against rainfall did not last long and they [emergency shelters] had no protection against winds or water (they flooded very often). Almost all the participants in the focus groups described the conditions of the emergency shelter solution as ‘infrahuman’, ‘unbearable’, or simply ‘very bad’.”
Additionally, the evaluation could find no documentation that “the decision to fix two tarpaulins per family as the main emergency shelter support is based upon the capacities of the population to complete the shelter solution,” and that in most cases agencies “did not take into consideration the family size when delivering the solution”.
“It did not matter if there were three or nine in a family, they gave everybody the same,” said one interviewee.
Confirming reports from the time that emergency shelter was often sub-standard and did little to provide protection both from the elements and from crime, the evaluation found that “some of the Sphere shelter standards indicators were not initially met” and that “site planning and design of emergency and interim shelter did not enhance protection or reduce the risks of gender-based violence, including sexual exploitation and abuse, by integrating aspects such as family-size shelters or partitioning shelters.”
Cost Overruns, Lack of Flexibility Prevent Durable Solutions
The evaluation found that international agencies focused on T-shelters to the detriment of other more cost-effective and efficient solutions and that despite the early indication that the goals for T-shelter coverage were unrealistic, agencies were too rigid in their plans and were either unable or unwilling to change tack. NGOs preferred to focus on T-shelters as they were more visible than rental support or repairing homes.
According to the report, the provision of shelter was based more on supply than demand, i.e. what the beneficiaries needed. For instance, the estimated number of T-shelters to be built (125,000) was based not on a needs assessment but rather on what shelter agencies had pledged to provide. The evaluation notes that decisions were made by agencies “based on their previous know-how, supposed ease of implementation, outcome control, liability concerns and/or visibility,” but not the actual needs of those affected. This became especially important as the provision of T-shelters became increasingly expensive and slow.
The original total cost of the all T-shelters was $187 million, with a time frame to complete distribution of just 12 months. In the end, the report finds that it will actually take over two years and cost $530 million and even then won’t cover the entire population’s needs, which have been revised upwards.
“The transitional shelter strategy could have been revised when it became obvious that goals and deadlines would not be met, resulting in a more comprehensive longer-term transitional shelter or permanent housing approach for part of the targeted population (for instance, a greater involvement in host families’ support and rental support could have lessened the burden to deliver transitional solutions), but shelter agencies’ programmes were not flexible enough, often because of their funding commitments, or could not easily be adapted on the field,” the report states.
The Shelter Cluster often left complimentary efforts such as rubble removal to other clusters and agencies, without realistically looking at their capabilities. Agencies involved in the Shelter Cluster alleged that donors were reluctant to fund rubble removal programs or had already earmarked funding for T-shelters. Additionally, the team found “agencies were also reluctant to spend their privately-raised funds on rubble clearance, for different reasons” including the lack of visibility of rubble removal efforts as compared to T-shelter construction.
As delays mounted, the T-shelters’ “added value progressively reduced, losing relevance and even acceptance with the local authorities and the affected population.” It is of little surprise then that in their consultations with local populations, the evaluation team found that “although they value the benefits of having been upgraded from the E-Shelter [Emergency Shelter], they feel the transitional solution does not meet their family needs.”
As international agencies stubbornly clung to the T-shelter, other more sustainable plans such as housing repair and rental support that may not have brought visibility to their organization’s work or were deemed too risky were side tracked.
While housing repairs were delayed, many Haitians have already moved back into severely damaged homes. A USAID-sponsored study found that over one million people were living in “extremely dangerous” houses, those marked yellow or red. The result is that nearly two years after the earthquake, durable solutions are still lacking.
“[T]he fact is that as of the end of 2010 (and may we say, up to date) there was no clear roadmap on what to do for permanent housing in the urban setting, no model or process had been outlined, no vision or guidance was in view, and most shelter agencies did not evaluate their real capacity to engage in housing repair,” the report states.