Community Perceptions and Behaviour: Results from a Field Survey

An important part of this study was a field survey of affected communities in the three most severely affected provinces. The motivation for this survey was to understand behavioural adjustments of families affected by the tsunami, so as to be able to draw lessons of wider relevance for policy formulation for natural disaster management.

Primary data was obtained from a sample survey of 262 displaced and non-displaced individuals, 37 visitors (tourists), and seven NGOs. The survey was designed to gain insights into aspects of the disaster management process as seen from the viewpoint of affected communities, tourists, and non-governmental agents such as NGOs. In particular, we were interested in obtaining community perceptions regarding the effectiveness of aid in terms of availability, distribution, speed, and targeting among various groups of occupations, and whether assistance may have adversely affected incentives for self reliance by creating an “aid-dependency” syndrome.

7.1 Survey Methodology

The survey was conducted in June 2006. It covered the three provinces severely affected by the tsunami and included 296 respondents. There were eight individual sample categories: Diversified Households’ (H) Fishers (F), Entrepreneurs (dependent mainly on tourism) (E), Labourers (L), Beach Vendors (B), Tourists (T), and Moken (an ethnic minority tribe of ‘sea gypsies’ who comprise a small fishing community in Phang Nga) (M), as well as seven NGOs from Phang Nga province. The sample distribution by province and category is given in Table 15 [ PDF 71.6KB | 1 pages ]. Sample size in each province was roughly proportional to the loss of life in each province, taken as an indicator of the severity of the tsunami impact. In each category, sample size was determined to obtain a reasonable size and cross section conditional on the overall budget for the survey, and the group sample sizes do not strictly correspond to their proportions in the population.

The families included in the Diversified Households group had family members engaged in a variety of occupations, including general service providers, fishers, housewives, and traders. Thus, while some of these household members follow occupations similar to those of the main income earners in the various occupational categories, the households as a whole have more diversified income sources. Twenty-three fishery households (Fishers) were randomly selected from Phang Nga, eight from Phuket, and five from Krabi. Those in the Entrepreneur category were mainly small business operators involved in tourism-related businesses such as restaurants, shop houses, and guest houses. The Labourer category includes those who were employed in hotels, restaurants, shop houses, department stores, and local administration. Beach Vendors provide various services to tourists on the beach, such as renting out beach umbrellas and chairs and selling food. The activities of NGOs have been concentrated primarily in Phang Nga, where the need for assistance was deemed to be strongest. The sample was drawn entirely from that province. We wanted to ensure that the NGOs could provide their own perspective to balance that provided by community groups. The Moken are fisher folk who also provide construction labour and engage in petty trade. They were included as a specific category in the survey to assess their perceptions as an ethnic minority group regarding any bias or discrimination in the provision of assistance.

Overall the sample was reasonably gender balanced (52 per cent males, 48 per cent females) but there is considerable variation among different groups. There were significantly more male respondents than females in the Fisher and Labourer categories (81 per cent and 58 per cent respectively) while there were more females within the Entrepreneur (64 per cent) and the Moken categories (56 per cent). In other categories (Diversified Households and Beach Vendors) the gender numbers were approximately equal.

Half of the surveyed Entrepreneurs and around a third of Diversified Households and Beach Vendors were new migrants to these areas, in contrast to nearly 80 per cent of the Fisherman and almost all of the Moken being natives of these areas.

7.2 Tsunami and Asset Losses

Information gained from survey respondents provided a household-level perspective on how the tsunami affected households (Figure 10 [ PDF 70.5KB | 1 pages ]). Some families in each group had lost family members, while most lost some assets. The way in which affected families have adjusted to post-tsunami life and coped with the many economic and non-economic challenges they have had to face have been conditioned by these two, key, traumatic driving forces: loss of family members and asset destruction. We will relate the adjustment and coping strategies of families to these two driving forces. Respondents were asked to approximate the damage on their property as a percentage of their total assets. In relative terms, Moken were the worst affected by asset losses, losing almost 80 per cent of their total assets. Laborers lost 40 per cent of their assets. In our sample, less than 20 per cent of Fisher families lost family members—the lowest rate among the six groups—but Fishers were hit hard by asset losses.

7.3 Adjustment and Coping with Post-Tsunami Reality

Table 16 [ PDF 70.5KB | 1 pages ] presents information about how families adjusted to the loss of incomes and livelihoods after the tsunami. Eighteen months after the tsunami, the incomes of the large majority of families (68 per cent) were below their pre-tsunami level, while a small minority (15 per cent) experienced an increase in income level.

But significant differences were found within the sample categories. Those who depended heavily on incomes from tourism were the hardest hit: 92 per cent of Beach Vendors and 86 per cent of Entrepreneurs experienced a decrease in income, followed by Fishers (72 per cent) and Diversified Households (69 per cent). Surprisingly, the Moken minority suffered the least from the reduction in income. Only a little over half of them experienced a decline in income level, 36 per cent of them were able to maintain the same level of income, and 11 per cent actually earned higher incomes. While 64 per cent of families belonging to the Labourer group had lower incomes, some 30 per cent reported that their incomes were higher.

More insights into the reasons for these different income level experiences can be gleaned from the ways in which families adjusted their income-earning activities after the tsunami. Overall, 40 per cent of the sample reported that they had changed their main occupation. The largest proportion of those who changed jobs was in the Labourer group of whom nearly half took up a new occupation. In contrast, only 30 per cent of Entrepreneurs and the Moken changed occupations.

A variety of factors to do with occupational skill-specificity, adjustment costs, and alternative job opportunities appear to have been important. Many Entrepreneurs probably felt that they were able to cope with a temporary income shortfall based on the expectation that tourist arrivals would recover before long. For the Moken, many of them were not only less dependent on markets for their incomes, but with limited skills, may have also had fewer opportunities to move into other jobs.

7.3.1 Consumption Smoothing and Access to Credit

A shock like the tsunami can be thought of as a temporary negative shock that reduces income for a period of time. Of course if significant asset damage also occurs, then there is a longer-term impact that reduces household wealth. However, the large scale assistance programmes generated a perception that asset damages would be more or less fully compensated through government and other donor aid. An expected response to income reductions, particularly when the decline is expected to be temporary, is an increase in borrowing as families attempt to smooth their consumption.

In the post-tsunami period, despite the availability of some aid and assistance, nearly half of all households across the different categories went into debt. However, though over half of the Moken families experienced an income fall, less than 20 per cent of them went into debt. In contrast, the number of families who borrowed was highest among the Fishers. It is possible that some families who wanted to borrow may have been unable to access credit markets and thus became liquidity-constrained. This is particularly likely to be the case for those who were asset poor and hence unable to borrow because of lack of collateral. Arguably, most Fishers expected that their assets and equipment would be replaced and hence expected that their income levels would recover after some time. This expectation may have been shared by many lenders. On the other hand, many Moken families were more likely to be liquidity-constrained because they were probably not seen as good credit risks and hence were more likely to have been denied access to credit. Further research is required, however, before any firm conclusions can be reached.

7.3.2 Determinants of Loan Demand

We further explored the demand for credit within households using a model where the demand for credit (amount of borrowing) was considered to depend on the following factors: the pre-tsunami monthly income (YB) multiplied by the number of months of unemployment (to proxy the income gap needed to maintain the pre-tsunami expenditure level), the cost of borrowing or the interest rate, the loan purpose (business or non-business), the availability of collateral, and the total amount of financial aid (TOTAL-AID) obtained from direct aid, relatives and friends. The detailed results are given in equation (1) in the Appendix.

The results indicate that credit demand is positively related to direct aid amount, availability of collateral, and whether the loan is for business purposes, and (as expected) negatively related to the interest rate. It is interesting that the total cash aid received complements rather than substitutes for total credit demand. The significance of the availability of collateral as a determinant of credit demand indicates that tsunami victims who have suffered loss of assets, and thus loss of adequate collateral, are likely to be liquidity-constrained. Poorer households with no collateral are also likely to be forced into the informal credit market where interest rates are higher. Such high-interest credit raises the cost of credit and thereby further reduces the capacity of the asset poor households to use the credit market for consumption smoothing. In this context, effective microfinance programmes can play an important role by providing credit to restart microenterprises to those who cannot afford the high interest loans.

Financial assistance from friends and relatives—from social networks—can be very important for cash-strapped tsunami victims experiencing income shortfalls. In our survey assistance from friends and relatives accounted for about a third of total financial assistance to households (Figure 11 [ PDF 91KB | 1 pages ]). However, this was not evenly spread across the different household categories.

Beach Vendors received the bulk of their financial assistance from relatives and friends, while Diversified Households and Labourer households received about a third of their total assistance from relatives and friends. On the other hand, outside assistance was by far the most important source for Moken and Entrepreneur households. In the case of Entrepreneurs, it is possible that this low proportion of financial assistance from relatives and friends was related to several factors: their need for handouts from friends and relatives to meet living expenses may have been lower; they had better access to savings; and credit markets and outside agencies, including the government, may have provided large scale assistance to restore business assets destroyed by the tsunami.

The Moken community was in a very different situation. If families and friends were also affected by the disaster, or if their friends and relatives were quite poor anyway, their capacity to help each other would be constrained. In this context, being a migrant may have been an advantage because it meant some chance of assistance from relatives and friends living at a distance and unaffected by the tsunami. In fact, after the disaster, the Moken community seems to have recognised the value of being more integrated with the broader Thai society. Moken people have obtained Thai citizenship and thereby have better access to formal education for their family.

Households whose livelihoods have been badly hit by a disaster can seek alternative employment, provided such employment opportunities exist or emerge after the disaster. As Table 16 showed, a significant number of households changed occupations. In Table 17 [ PDF 71.2KB | 1 pages ], we examine the degree to which changes in occupations enabled maintenance of income levels within different household groups. Most households were forced to accept lower incomes whether or not they remained in the same occupation. This does not mean that changing jobs was of no benefit; those who changed jobs did so because the alternative would have been an even lower level of income or unemployment, and the cost of changing jobs was more than compensated for by the expected earnings in the new occupation.

For many people employed in the tourism industry, or who depended on it, there was considerable pressure to change jobs. As described earlier, the tourism industry was severely affected by the tsunami. As the number of tourists declined, tourist-related businesses were adversely affected and many hotels and restaurants closed down or scaled down their operations. This led to job losses. In some cases, workers accepted lower wages in order to maintain their jobs. Those who lost their jobs needed to find new ways to earn a living. But if their skills were industry-specific it was difficult to find new occupations that could even maintain the previous income, let alone provide a higher income. Hence, many people who remained in their previous occupation were compelled to accept a lower income and only a small percentage could maintain their previous income level.

As tourism began to recover, wages improved for those employed in the hotel business. There were shortages in specific skilled labour categories; some of the skilled workers had been killed by the tsunami. Many Labourer households who were employed in the tourism industry and remained in their occupation experienced an increase in income by the time of our survey in mid-2006. But the recovery in tourism was not yet sufficient to restore business for Beach Vendors to previous levels. Over 90 per cent of those who remained in that occupation continued to have a lower income even in mid-2006. There was a similar story in the case of Fishers whose income declined for several reasons. The tsunami destroyed their boats and fishing nets, limiting their capability to fish. In the first few months after the tsunami, fish consumption in Thailand fell because of public fear that the fish were feeding off corpses in the sea. Furthermore, the oil price shock in 2005 raised the price of diesel, a major cost component in the fishing industry. Thus, even when Fishers were given fishing gear, these factors constrained their capacity to fully recover their earlier income levels quickly.

Figure 12 [ PDF 105.4KB | 1 pages ] shows the proportion of households in each group that have engaged in borrowing and the proportion of households that did not change their pre-tsunami occupation. Among those who remained in the same occupation, the percentage of families with debts was high, ranging from less than 25 per cent among the Moken people to more than 50 per cent among Beach Vendors, Entrepreneurs, Fishers, and Diversified Households. The Moken families, despite their income reduction, incurred lower debts (possibly because they could not access credit) and experienced serious distress. The other surveyed households, in general, had better access to credit, thus easing the hardship despite their inability to change occupations. Furthermore, many of those who did not change occupations were prepared to remain in their occupations in expectation of the recovery of the tourism and fishery industries. In this case, access to credit enabled them to tide over the expected temporary shortfall in income.

To sum up, whether or not they changed occupations, households affected by the tsunami experienced significant income losses. But most people saw the tsunami as a temporary shock and expected the situation would return to more to normal levels after some time. Because of this, many felt that it might be best to maintain the same occupation while waiting for the recovery of the relevant industries. A new occupation can be useful as a temporary job to make ends meet in the short term but it was unlikely to be a better occupation in the long run.¹⁴

This raises the issue of the effectiveness of the occupational training for the tsunami victims which was provided by many agencies. These job training programs may not actually have been of much help. In the first place, people are less likely to look for new occupations if the current bad situation is seen as only as a temporary one. In any case, for people to be attracted to new training programs they need to be able to see a market for the new skills that will provide adequate compensation for the costs of long-term occupational change. Our survey responses indicated that many people were not convinced that the benefits from such training were sufficiently large to compensate for the costs. This may hold a lesson of general relevance for job training programs in disaster situations: If the disaster is seen as only a temporary shock, there may be little incentive for people to undertake substantial training to prepare for alternative long-term occupations.

7.4 Impact on Mental Health: Post-Traumatic Stress Disorder, Anxiety, and Depression

A group of Thai and international medical researchers associated with the Thailand Post- Tsunami Mental Health Study Group have documented the significant mental health problems that affected many tsunami victims. Their work was based on surveys conducted in February 2005, only a few weeks after the tsunami, and a follow up survey in September 2005 (van Griensven et al., 2006). They found a significant incidence of anxiety, depression, and post-traumatic stress disorder (PTSD) eight weeks after the disaster. Nine months after the disaster, these problems had diminished but had not entirely disappeared.

It is not our intention to examine the issue of mental health in detail in this study. Nevertheless, because information on the mental health problems of our survey respondents might be useful, we included several questions that enabled us to construct an index of mental health. The index was constructed on the following basis: If a person is able to carry on with their normal way of life as if nothing has happened, a score of zero is given. If a person accepts their misfortune, but acts with resolve and copes well with the crisis, a score of one hundred is given. The score is raised to two hundred if they are still in a state of acute sadness, to three hundred if they suffer from hallucinations, and to four hundred if they are paranoid.¹⁵ Hallucination and paranoia are symptoms of post-traumatic disorder. Admittedly, the index constructed in the survey is not comprehensive like the Harvard Trauma Questionnaire and the Hopkins Checklist-25 to detect symptoms of anxiety and depression. Further, respondents may not reveal the true state of mental stress suffered depression and PTSD eighteen months after the tsunami. For a group, a mental stress index computed as an average equal to one hundred indicates that the group has resolved its depression and can cope with the stress in a stable manner. If the index is closer to two hundred, it suggests that the group members are still quite saddened by the traumatic experience. If the group index is higher it indicates that hallucinations and paranoia are common, suggesting acute PTSD.

Table 18 [ PDF 73.7KB | 1 pages ] presents the group averages of our mental stress index and some other indicators associated with mental stresses or concerns. The group averages suggest that, on average, people are coping with the effects of the tsunami: They are still affected—“sad”—but are able to accept the tragedy and cope with it. Even the Moken families, who have a higher stress index (144), are not in an acute state of stress or paranoia. However, some individuals within each group do suffer from acute mental problems.

People who have lost their assets and properties experienced higher mental stress. Moreover, in all groups, anxiety and fear of another tsunami continued to be present among around half of the respondents. In each group some people reported knowing someone who had attempted suicide, but the proportion was somewhat lower among Fishers and the Moken people. Only around half the respondents reported participating in a tsunami drill; it is not clear if this helped alleviate anxiety and stress.

As mentioned in Section 6.7, mental stress is aggravated further if social tensions have increased within the community due to perceptions of inequitable distribution of assistance. Respondents were asked about community unity after the tsunami, with a score of one if it had strengthened, 0.5 if it was unchanged, and zero if it had deteriorated. If the average group score was below 0.5, it suggests that community cohesion had deteriorated after the tsunami. The group average was less than 0.5 in all cases indicating a widespread view— though the extent of this differed between groups—that community cohesion had deteriorated rather than strengthened after the tsunami.

We used an Ordered Probit regression to explore the relationship between mental health and a number of likely explanatory variables. The index (MENTAL) was constructed as earlier described, and this was regressed on the percentage of total assets destroyed by the tsunami (LOSSASSET), expectations of a future tsunami, interactions with other victims who attempted to commit suicide (SUICIDE), age, the number of family members lost in the disaster (LOSSFAMILY), consultation or visits provided by health officials (CONSULT), and perception of community unity (UNITY).

The results (see equation (2) in the Appendix) suggest that economic loss is a key factor significantly affecting mental health. This finding is consistent with results reported by van Griensven et al. (2006), who found that the loss of livelihood was independently and significantly associated with symptoms of all three mental health outcomes (PTSD, anxiety, and depression). Loss of family members and expectations of future tsunamis, too, adversely affected mental health. On the other hand, a perception that community unity has been strengthened helped reduce mental distress, as did visits or consultations by health officials. Age and knowing someone who had attempted suicide were not significant. These results have obvious policy implications, highlighting the importance of adequate and equitable financial assistance, and the value of visits by health officers to provide counselling and advice.

7.5 The Tsunami and Tourist Attitudes

Given the importance of tourism to the regional economy, we analysed responses obtained in our survey from a small sample (37) of tourists visiting the three provinces to explore their views and attitudes to gain some insights into ways in which the industry could recover rapidly. The sample was evenly distributed between new visitors and previous visitors who had returned to the tsunami-affected beaches. In Table 19 [ PDF 86.3KB | 1 pages ] we present results from our survey.

The views expressed by the tourists give grounds to be optimistic about the recovery of the tourism industry. Some two-thirds of the respondents did not expect another tsunami to hit the area within the next few years. In addition, and importantly, they all agreed that the early warning system installed by the government can improve their level of safety. Despite the fact that 22 per cent of the tourists surveyed had known someone who had died in the tsunami, these visitors had, nevertheless, opted to return to the Andaman area. During the first quarter of 2005, unlike Western visitors, Asian visitors reportedly tended to stay away from the affected areas because they felt uneasy about the tsunami ghosts. In our sample, more than eighty per cent of the visitors said that were not afraid of tsunami ghosts and spirits. A slight majority (57 per cent) agreed that the media had scared away the tourists, and two-thirds of visitors still saw some traces of the tsunami’s destruction.

The respondents rated the rehabilitation performance of the Thai government in three areas: infrastructure reconstruction, environment restoration, and assistance given to the tsunami victims. The overall rating on a scale of 1 to 5 across all three areas was relatively similar and satisfactory: 3.32 (66 per cent) for infrastructure rehabilitation, 3.38 (68 per cent) for environment restoration, and 3.68 (74 per cent) for victim assistance.

We also asked a hypothetical question to gauge the impact of a tsunami in terms of visitor perceptions of safety (“fear”): How many months would it take the respondents to return to the areas if another tsunami were to strike in the future? If visitors have no fear, we can presume they would return without hesitation when they have the opportunity. But others may be deterred from returning by several factors—if they greatly fear the occurrence of another tsunami, it may take time before this fear subsides; a grim picture painted by the media; a poor reconstruction effort by the government; fear of tsunami spirits and ghosts—all these factors may lead them to go elsewhere. While recognising that the sample included only visitors who had already returned, and hence was biased, we used a multiple regression model to get some indication of the extent to which these factors may influence tourists to delay returning to the region, taking the measure of “fear” as a proxy.

The results suggested that the painting of a grim picture by the media is very likely to have a significant deterrent effect, interacting with the perception of tourists on the rebuilding effort of the government (see the detailed results in equation (3) in the Appendix). The fear of tsunami ghosts, interacting with the number of previous visits, was the most significant variable in the equation, more important than the variables representing expectations of a future tsunami and the positive effects of successful government efforts in rebuilding.

Even if tourists return, a critical issue for the tourism industry is the duration of stay. We ran another regression, using the number of days visitors would stay in the area as the dependent variable, hypothesized to depend on the fear factor, the media scare variable, and an Asian dummy variable, given that Asian visitors tend to have different behavioural characteristics to other visitors. The results (for details, see equation (4) in the Appendix) suggest that fear, as expected, lowers expected duration of stay. Further, any media scare is particularly important as a negative factor in the case of Asian tourists. This finding—if confirmed by more studies of a more representative sample—would indicate that efforts to promote tourism through discount holiday packages and marketing promotions may not work that well for Asian markets. On the other hand, if media coverage is favourable, highlighting the successful rehabilitation of the areas and the tourism industry overall, this may have a more positive impact on Asian tourism.

7.6 Assistance Assessment

7.6.1 Quality, Level, and Effectiveness of Assistance

A major aim of the field survey was to obtain the views of the tsunami victims in each of our categories about the programme of aid and assistance for relief, reconstruction, and recovery. Respondents were asked to evaluate the overall assistance from various organizations on a scale of 1 to 4, with a score of zero when there was no help at all, 1 for a low level of assistance, 2 for medium assistance, 3 for high assistance, and 4 for the highest level of help.

The survey results indicated that respondents in all categories received help and assistance from government, foundations, and private corporations. In many cases, the role of government assistance is complementary to corporations and charity foundations. The government and other organizations sometimes pooled their resources in, for example, providing housing for families whose houses had been destroyed. Similarly, private corporations contributed to financing housing construction done by army units. None of the survey respondents reported any systematic bias in the provision of government assistance.

However, the overall level of satisfaction was not very high (Figure 13 [ PDF 105.6KB | 1 pages ]). Still, this response must be treated with some caution—it is possible that respondents may have rated the overall assistance not very highly in hopes to get more assistance in the future. Moreover, there was substantial variation among the groups in their assessments of the various donor agencies. While Diversified Households, Entrepreneurs, Labourers, and Beach Vendors ranked assistance from the government equal to or higher than the assistance from the private sector and foundations, Fishers and Moken people rated assistance from foundations and private corporations as superior.

We asked a series of questions about their cash needs, the amount received, the speed and adequacy (Figure 14 [ PDF 106KB | 1 pages ]). Again, the responses have to be treated with caution because respondents are likely to have felt that their answers might affect the amount that they would get in the future, so they may have tended to overestimate their demand for cash aid while underestimating the amount they have received.

Respondents rated the amount of cash grants they had received on the following scale: 0 for no grant, 1 for little, 2 for moderate, 3 for large (high), and 4 for very high. In addition, they were asked to rate their need for future cash grants on the same scale. While there were again differences between the categories, the overall responses indicated that what they had received was low to moderate, and that low to moderate levels of cash assistance were still needed.

As for the speed of cash disbursement, respondents who did not get any cash were assigned a score of zero, 1 if it took more than a month, 2 if two weeks, 3 if within one week, and 4 if they got the cash immediately. Most people got cash aid within a month. The average rating for speed of cash delivery was 1.5, implying that cash aid, on average, was received within about three weeks of the tsunami, but in the case of some groups it was quicker (one to two weeks)—including Fishers and Moken people—but for others, including Beach Vendors and Labourer households, it was closer to a month.

For the question as to whether they received a sufficient amount of cash aid, if the respondent answered no, the score given was zero, and if yes, the score given was one. (Note the difference in the scale for this question when interpreting the height of the column in the Figure). The score for this sufficiency question was very low, ranging from 0.08 (Labourers) to 0.19 (Moken). Again, these scores may have been highly influenced by the perception that responses may influence future assistance levels.

As described earlier, another aspect of assistance was provision of job training. According to the survey responses, job training assistance came relatively late, about seven to twelve months after the tsunami.

In Figure 15 [ PDF 89.7KB | 1 pages ] we present the survey responses across groups to questions on areas of need for future assistance. Respondents ranked each category of demand on a scale of zero to four. Leaving aside the question on business assistance, which was only directed to respondents in the Entrepreneur group, the type of assistance that respondents ranked as being most important was cash grants, followed by job training, housing, and health care.¹⁶

7.6.2 “Build Back Better”

In all the tsunami-damaged countries, reconstruction activities have been conducted under the slogan “build back better”. The basic idea has been that new construction would not simply replace what existed previously, but would be superior in quality. We investigated community perceptions of the effectiveness of this approach to reconstruction by focusing on a key component of the reconstruction assistance effort: the reconstruction of damaged or destroyed houses. Respondents who had received housing assistance were asked to indicate whether the quality of their newly constructed house was better than their old house.

There have been reports both in Thailand and in other tsunami-affected countries of many instances where communities have not been happy with the new houses. Arguably, people who have been greatly traumatised by the experience of a disaster can react to a new home in different ways. In their mentally depressed state they may view the new house as being an inferior replacement even if it is in fact a superior structure. Alternatively, they may rate it as better because they have lowered their expectations due to reduced circumstances, with unemployment, debt, and so on as pressing factors.

We investigated this using a model with a dummy dependent variable, Build Back Better (BBB), which was assigned the value of one if the new house is perceived as a better house than the old one (0 otherwise). (See the details in equation (5) in the Appendix.) The Probit regression result suggests that the economic circumstances of households may be related to how they view their new houses. Households which have suffered serious economic losses, have experienced a longer period of unemployment, or are in pressing need of more cash assistance (a proxy for being liquidity-constrained) are more likely to rate their new house as better.

Psychological factors (mental distress and loss of family members) rank lower than economic factors. It must be recognised that there is a co-linearity problem between the number of family members lost and mental health deterioration, which may lead to the coefficient of psychological factors becoming statistically non-significant. While our results are not conclusive, it does appear that those who lost almost everything tend to appreciate the value of their new houses, while those who did not lose much appear to have higher expectations of the quality of the houses offered to them. Those who suffered relatively smaller losses tended to deny the notion of build back better.

7.6.3 Role of NGOs

We also obtained some information on community perceptions of the role of NGOs, and, through interviews with seven NGOs, obtained some information from the NGOs’ perspectives.

The NGOs were involved in a range of assistance activities. Some of them have provided livelihood equipment and job training, initiated health consciousness campaigns, and set up organic farms and child care centres. They also provided education-related assistance such as providing English teaching, giving student scholarships to tsunami-affected children, and building community libraries.

There have been problems. Many of the NGOs interviewed expressed the view that gaining cooperation from villagers can be a problem. There tends to be a perception in the community that the NGOs will only be there in the short-term, and thus cannot be relied on for longer-term assistance. ¹⁷ This makes people reluctant to contribute to or participate in their programs. Some NGOs admitted that lack of coordination among them has produced overlap, leading to inequitable and inefficient delivery of assistance. The result has been to produce a sense of unfair treatment among recipients, and the deterioration of social cohesion among villagers. On the other hand, many NGOs encountered avaricious aid recipients who resorted to deception with fake claims of damages and injury—sometimes even using corpses that belonged to other families—to claim compensation and assistance. With assistance focused on material and cash handouts, some children were reportedly disappointed that they were not orphaned because they received less assistance from the NGOs.

7.7 Aid Dependency?

In the literature and debates on aid, a frequently expressed concern is that aid recipients may develop “aid dependency” which creates an attitude that constrains them from taking private initiatives to move beyond aid recipient status. While eighteen months after the disaster may be a relatively short period to assess this, we undertook some preliminary analysis of survey responses to see if any sign of such a syndrome is apparent. Some indications of the development of an “aid dependency” syndrome may be discerned if people continue to rate their assistance needs high even when they have had substantial assistance (and, therefore, should now have lower assistance needs). When we asked the respondents if they still needed additional cash aid in 2007, those who suffered greatly from asset destruction and income reduction were more likely to reply “yes”. But those who have received a new house and a secured land title are unlikely to require more cash, and should need to depend less on financial aid. We explored whether these types of propositions are supported by the survey responses.

An index was constructed to measure the degree of cash dependency by taking the ratio of desired cash in 2007 to the cash aid respondents had received in 2006. If the log value of the ratio of need for cash (nmcash) to the cash received (recash) is negative, households’ expressed need for cash aid was considered lower. If the value was positive, the expressed need was considered higher, suggesting a tendency towards growing aid dependency. As seen in Figure 16 [ PDF 70.9KB | 1 pages ], the kernel density distribution is tilted to the positive region, indicating that there is evidence supporting the aid dependency view. However, this needs further analysis to examine why the respondents continue to ask for more cash aid.

We explored this further by undertaking a multivariate analysis, defining a dependent variable MORECASH as the difference between cash “needed” in 2007 and the sufficiency of cash previously received. A Tobit model was used because respondents are likely to overstate their amount of cash aid needed, hoping to receive more cash, and are likely to understate the amount of aid they received. Hence we used a censored model with extreme values, -3 and 3, resulting in three observations left censored and ten observations right censored.

The results (equation (6) in the Appendix) show that those who believed they had received adequate cash (SUFCASH) tend not to express a need for more financial assistance. Also, it is interesting to note that those who have borrowed tend not to seek more money. A possible explanation is that the fact that they have borrowed indicates they have access to credit markets, hence they are not highly credit-constrained and do not seek more handouts from donors. Looking at the variables that are statistically significant and positive, only those who have lost their livelihoods through mental depression and suffered a large income loss after the tsunami appear likely to continue to demand more cash aid.

In sum, after taking into account these factors, we conclude that there is no strong evidence to support the aid-dependency view. The finding that those who are able to adjust through borrowing tend not to ask for continuing handouts suggests that once the difficulties in the immediate aftermath of the disaster are overcome, tsunami victims will not become aid dependent if they have better access to credit markets. In addition, employment-generating programs—which would enhance their capacity to both borrow and repay—are likely to give them a better chance to lead their normal lives without having to depend on aid in the future.

7.8 Other Impacts of Assistance

There are several other aspects related to aid and assistance that we have not explored. We observed, for example, that there were concerns that the cultural heritage of the minority Moken people might be lost due to the outpouring of assistance, which some fear has eroded their traditional way of life. Moken children, some argue, no longer care about their language and culture, and money and material belongings have become the dominant objectives in their lives. They blame this on the fact that their rebuilt community is located near the road linked to city areas. Donors must be sensitive to such unintended impacts of large-scale assistance and reconstruction activities.

Download this Discussion Paper [ PDF 440.1KB| 45 pages ].

Comment(s)

There are [0] comment(s) for this entry. Post a comment.