mechanisms behind the process of decline of sasi: What makes sasi strong or weak, and why
do some villages want to revitalize sasi? For all the villages, every aspect of sasi is described,
to a large extent, in the respondents’ terms.
2.5 Comments on the Methods: Gaps and Limitations
The inventory provides 100% coverage of villages on the three Lease Islands (Saparua, Haruku,
Nusa Laut). However, villages in Seram and Ambon were not randomly selected; therefore
the data from these islands cannot be assumed to reflect the average condition of the sasi
institution there. In fact, we know that, because we sought out villages with marine sasi on
Ambon, the percentage of villages in the Ambon data having sasi over-represents the actual
occurrence of sasi on this island. The number of villages sampled on Seram (5% of all coastal
desas) is too small to provide a picture of sasi on that large island.
Because knowledge about sasi varies greatly among village respondents, we tried to interview
at least three key informants in each village when doing the inventory. In a few cases, the
desired key informants were not available and information came from only one or two persons.
Also, information from different respondents was sometimes contradictory or incomplete. In
developing the database, a weight of evidence approach was used to assign variable values
in such cases. Sources of additional information included Ambon-based researchers and NGO
field workers as well as the surveys and interviews conducted in 28 of the villages in the
course of this project.
Since each village institution is different, the “institution” being evaluated is really a broad
range of institutional arrangements. This makes the results more variable and difficult to
interpret than results from one single project site.
In order to reduce some of the complexity we found in the field, all case studies were performed
in Christian villages. The subsequent analysis of sasi is, therefore, most relevant to those
forms of sasi practiced in Christian-dominated areas. We have noted certain differences
between Christian and Muslim institutional arrangements in the course of our inventory. The
analysis of Muslim-style sasi is a topic for future research.
Interviews of fishers as well as interviews in the case study sites were conducted with the
consent and under the guidance of the village head of each village. Because of this, the samples
were not random, although, for the case study sites, efforts were made to engage people in
the street in informal discussions and interviews. For performance survey interviews in
particular, the village head selected fishing families to participate in the survey. It is, therefore,
possible that respondents represented a middle-class range of fishing families rather than the
entire spectrum available in each village.
The ladder survey instrument was adapted from the methodology used to evaluate CBRM projects
in the Philippines. When used to evaluate the impact of a project, time lines are short and distinct.
Respondents indicate their perception of conditions before and after a project is carried out.
However, with sasi, there is no “before” or “after”. We used 15 years as a time boundary because
it was roughly 15 years ago that village governments were re-organized, leaving traditional
management institutions without legal status. With such a long time line, perceptions of trends
through time can be expected to vary, depending on the age of the respondent.
Methods 25