Poverty in Seychelles: Policy Digest Christophe Muller (United Nations Development Programme Consultant, July 2012) This document reports on the estimation of a poverty line and a poverty profile for Seychelles in 2006. The used notion of poverty is correspond to the opinions of the consistent respondents about total basic needs in the Seychelles population. The estimated poverty line is worth 13554 Rupees per adult-equivalent per year. The poverty rate is estimated at 17 percent of the population. Poverty is found higher in households led by unemployed heads, or by female or little educated heads. Other categories of households especially affected by poverty are the large families and fishermen families. A few detailed tables are reported below. The estimates are based on 1161 observations. The sampling standard errors are in parentheses. The following poverty estimators are presented in the tables. P 0 = poverty rate, P 1 = Poverty gap, P 2 = Poverty severity, W = Watts poverty index.
Table 1: National Poverty Estimates Poverty estimates (in %) P0 16.96 (1.40) P1 4.62 (0.48) P2 1.88 (0.27) Watts 6.19 (0.73) The estimated poverty rate is about 17 percent of the population. That is: there are 17 percent of poor persons in Seychelles, based on this methodology and survey data. This figure corresponds to a broad notion of poverty based on the opinions of Seychelles households on subsistence minima expressed in terms of total consumption expenditure, including housing expenses. This is not comparable with the poverty measures based on nutrient minima that are used for some extremely poor countries, which is not the case of Seychelles. Indeed, poverty understood as suffering of hunger would yield almost nil estimates in a country like Seychelles, and therefore would be a useless statistics for managing social programs. The obtained percentage of the poor is also appropriate for running social policies in this country in that the corresponding target population in neither negligible nor beyond the capacity of the public budget. In particular, the found extent of poverty justifies the 2
efforts devoted to building a modern social system through social security and social welfare agencies. Finally, this poverty estimates compares well with the magnitude of other destitution figures obtained from analysing the LCS data and that will be presented in another report. The estimated poverty gap is at 4.62 percent, the poverty severity measure at 1.88 percent and the Watts index at 6.19 percent. Table 2: Poverty by household size (%) Household size P0 P1 P2 Watts 1 2.0 (1.17) 0. 7 (0.48) 0. 4 (0.30) 1.0 (.0076) 2 5.1 (1.60) 3 9.0 (18.28) 4 11.4 (2.10) 1.5 (0.56) 2.1 (0.52).035 (0.79) 0.6 (0.27) 0.7 (.0027) 1.5 (0.42) 2.0 (0.76) 2.6 (0.76 ) 4.8 (.0115) 5 15.3 (2.70) 3.4 (0.77) 1.2 (0.38) 4.3 (1.08) 6+ 32.3 (3.73) 9.2 (1.37) 3.9 (0.82) 12.5 (2.17) Whatever the poverty indicator considered, the estimated poverty is clearly increasing with the household size. This suggests that poverty alleviation policies should particularly support large families, notably families with many children and dependents. As a matter for fact, considering the very high poverty indicators that correspond to 3
them, families with six or more members should be viewed as a priority for social policy. Table 3: Poverty by gender of household head (%) Gender of Head Male 13.6 (2.03) P0 P1 P2 Watts 2.8 (0.51) 1.0 (0.25) 3.6 ( 0.72) Female 19.3 (1.92) 5.9 (0.73) 2.5 (0.43) 8.0 (1.14) Poverty is notably higher in female-headed households than in male-headed households. This is a common feature found in most of the economic literature on poverty. However, the situation is different in Seychelles. In most poor countries of the world, this observed feature is much due to the fact that households led by female heads are mostly led by widows, and widows are often poor since they are inactive and they is no adult male bread-earner in the family. In contrast, in Seychelles a majority of households is led by female heads. It is therefore likely that their higher poverty levels is more caused by other factors in Seychelles society, such as a non-negligible proportion of lone mothers raising their children on their own and with difficulty for having a productive activity because of this burden; or gender wage segregation on the labour market. 4
Table 4: Poverty by District (%) District P0 P1 P2 Watts Takamaka 22.5 (8.0) 5.7 (2.02) 1.8 (0.75) 6.9 (2.51) Anse Aux Pins 15.8 (6.66) 3.5 (1.62) 1.1 (0.65) 4.3 (2.06) Anse Boileau 29.0 (7.90) 7.2 (2.53) 2.8 (1.29) 9.4 (3.54) Au Cap 18.3 (6.25) 3.8 (1.85) 1.4 (0.76) 4.8 (2.37) Anse Etoile 14.7 (5.91) 3.2 (1.48) 1.0 (0.59) 3.9 (1.89) Anse Royale 21.0 (9.11) 2.8 (1.39) 0.7 (0.59) 3.4 (1.79) Bel Air 3.5 (3.43) 1.9 (1.89) 1.1 (1.05) 2.8 (2.76) Baie Lazare 41.7 (9.29) 11.9 (4.37) 6.2 (3.22) 18.1 (8.03) Belombre 12.5 (5.93) 2.7 (1.38) 0.8 (0.43) 3.2 (1.65) Baie Saint Anne 6.5 (3.95) 1.8 (1.31) 0.60 (0.50) 2.2 (1.64) Beau Vallon 14.7 (6.07) 4.8 (2.41) 2.1 (1.26) 6.8 (3.43) Cascade 25.0 (7.29) 8.21 (3.06) 3.8 (2.17) 12.0 (5.58) English River 11.2 (6.30) 2.2 (1.58) 0.6 (0.53) 2.6 (1.92) Glacis 10.8 (5.46) 3.2 (1.60) 1.2 (0.74) 4.1 (2.15) Grand Anse Mahe 14.6 (6.97) 5.6 (2.88) 2.4 (1.34) 7.4 (3.88) 5
Grand Anse Praslin La Digue 7.6 (4.53) 10.4 (6.19) 2.6 (1.96) 1.8 (1.02) 1.4 (1.06) 0.5 (0.33) 3.7 (2.82) 2.1 (1.22) Les Mamelles 6.25 (4.49) 0.6 (0.40) 0.05 (0.04) 0.6 (0.42) Mont Buxton 11.9 (6.20) 2.6 (1.61) 0.7 (0.47) 3.0 (1.90) Mont Fleuri 31.3 (9.38) 6.7 (2.52) 1.89 (0.88) 7.9 (3.08) Plaisance 9.0 (5.19) 3.16 (1.97) 1.2 (0.87) 4.1 (2.61) Port Glaud 17.6 (7.93) 7.03 (3.43) 3.5 (1.84) 10.1 (5.06) Pointe Larue 12.9 (5.13) 3.8 (2.90) 1.61 (1.10) 5.1 (2.93) Roche Caiman 40.9 (9.25) 14.6 (4.24) 7.1 (2.59) 20.8 (6.61) Saint Louis 9.47 (6.90) 3.3 (2.66) 1.2 (1.05) 4.05 (3.38) The large standard errors reported in parentheses imply that it is probably too bold to want to produce accurate estimates of poverty at district level with the used data. Indeed, many district poverty estimates are in fact not statistically different. However, the general picture of geographical poverty in Seychelles shown by these estimates seems to generally concord with accepted notions by NSB statisticians, with a few exceptions. Nonetheless, special care should be taken with some very high or very low poverty rates in that case, which should be confirmed by further local investigation. 6
Table 5: Poverty by Highest Level of Education of the household head Education Level P0 P1 P2 Watts No schooling 33.0 9.6 3.6 12.2. (7.60) (2.65) (1.30) (3.65) Primary 20.7 (2.39) 5.6 (0.83) 2.3 (0.48) 7.6 (1.29) Secondary 14.1 (2.19) 4.1 (0.87) 1.8 (0.55) 5.7 (1.44) Vocational and Polytechnic 12.0 (2.95) 2.9 (0.77) 0.9 (0.30) 3.5 (0.98) Pre University and University 0 (0) 0 (0) 0 (0) 0 (0) Poverty in Seychelles is a quickly decreasing function of the education level of the head, as in many countries in the world. No household has been observed as being poor with a head educated at university or pre-university levels. In contrast, even secondary or vocational education of the head is not a guarantee of escaping poverty. Household with heads without any schooling have particularly high degrees of poverty. One obvious conclusion of this table is that education should be one of the main way of fighting poverty in Seychelles. 7
Table 6: Poverty by Regions Region P0 P1 P2 Watts Central 16.4 4.55 1.83 6.03 (2.56) (0.89) (0.47) (1.28) East/South 18.7 (2.81) 4.92 (0.91) 1.94 (0.51) 6.54 (1.41) West 25.1 (4.16) 7.25 (1.63) 3.27 (1.05) 10.18 (2.72) North 13.2 (2.95) 3.47 (0.88) 1.25 (0.41) 4.40 (1.19) Praslin/La Digue 7.73 (2.69) 2.08 (0.90) 0.83 (0.42) 2.71 (1.22) The highest poverty levels are reached in the West region, while the lowest poverty occurs in Praslin and La Digue. Table 7: Poverty by Industry of Head of Household (%) Industry P0 P1 P2 Watts Unemployed and not stated 17.3 (2.43) Agricultural, Fishing and 28.2 Quarrying (9.68) Manufacturing, energy and water 15.1 and construction (4.99) 3.7 (0.63) 11.1 (5.40) 3.5 (1.37) 1.2 (0.27) 6.1 (3.97) 1.4 (0.67) 4.6 (0.83) 17.6 (10.04) 4.7 (1.93) Trade and repairs and transport 18.6 (5.98) Hotels and restaurants 15.9 (4.05) Other services 17.8 (8.00) Professional, scientific and 16.2 administration (2.62) 3.4 (1.47) 5.2 (1.56) 6.8 (3.54) 5.0 (1.02) 1.1 (0.70) 2.2 (0.79) 3.4 (2.17) 2.2 (0.62) 4.2 (2.0) 6.9 (2.18) 9.8 (5.58) 6.9 (1.64) Education and health 14.4 (4.65) 5.3 (1.86) 2.4 (0.98) 7.3 (2.70) 8
The poverty estimates by sector follow the aggregation pattern of the ISIC industry nomenclature. The industry sector with the highest poverty levels is that of fishing, to which has been grouped the small agricultural sector. Indeed, fishermen make the most destitute population in Seychelles. The ranking of the other sectors is ambiguous for two reasons. First, the sampling standard errors do not ensure that close poverty estimates for two different sectors are always significant. Second, the ranking of sectors depends on the preferred poverty indicator. On the whole, there does not seem to be a strong correlation between poverty and the industry sector of the head, apart from the fishing sector. Table 8: Poverty by major occupation of the household head (%) Occupation Category P0 P1 P2 Watts Managers, professionals and technician 7.00 (1.72) 2.06 (0.60) 0.83 (0.32) 2.73 (0.88) Clerical and service workers 19.3 (3.09) 6.10 (1.11) 2.62 (0.69) 8.32 (1.87) Manual workers 21.9 (2.82) 6.37 (1.11) 2.84 (0.68) 8.86 (1.80) Unemployed 28.0 (8.00) 5.96 (1.82) 1.62 (0.59) 7.00 (2.19) Other inactive 15.5 (3.20) 3.18 (0.83) 1.04 (0.38) 3.95 (1.12) The category most affected by poverty is that of households with unemployed heads, as it is often the case in other countries. Households led by manual workers, and to a smaller extent by clerical and service workers also suffer from a substantial amount of 9
poverty. In contrast, households whose head is a manager, a technician or a professional are rarely observed as poor. Beyond the hierarchies of remunerations by activities, a lesson from this table is that poverty alleviation policies should coordinate closely with labour market policies, in particular to provide jobs to unemployed heads of families. 10